Transconductance Ratio Research Articles

Carbon nanotubes field emission integrated triode amplifier array

Because the high frequency operation of a field emission triode amplifier is dictated by the cutoff frequency and not the electron transit time, a high ratio of transconductance, g m to the overlapping interelectrode capacitance, C g is the desired outcome. Consequently, to achieve high frequency performance of the CNT amplifier array in this study, C g was reduced by performing a dual-mask photolithography process to minimize the overlapping gate area, and, the insulating layer's thickness was increased. Moreover, wedge-shaped CNT emitter arrays are employed to increase emission sites, resulting in return higher g m. Both dc and ac performance of the amplifier were characterized. The triode amplifier array exhibited a high current of ∼ 0.32 mA (74 mA/cm 2), g m of ∼ 63 μS and voltage gain of ∼ 18 dB. Frequency response of the triode amplifier up to 20 kHz was also investigated. A theoretical cutoff frequency of > 70 MHz could be achieved with proper shielding of the test setup.

STUDY OF DUAL-MATERIAL GATE (DMG) FinFET USING THREE-DIMENSIONAL NUMERICAL SIMULATION

In this work, the novel characteristics of a FinFET with dual-material gate (DMG) are explored theoretically using a 3D numerical simulator and compared with those of a single material gate (SMG) FinFET in terms of threshold voltage roll off, drain induced barrier lowering (DIBL) and the ratio of transconductance (gm) to drain conductance (gd). Our studies show that the DMG structure achieves simultaneous suppression of short channel effects (SCEs), enhancement in carrier transport efficiency and transconductance. Also, these features can be controlled by engineering the work function and length of gate material.

Analog implementation of wavelet transform based on switched-current filter circuits

A general scheme for the analog implementations of wavelet transform using switched_current (SI) filter circuits is presented. SI circuits are well suited for this application since the dilation constant across different scales of the transform can be precisely implemented and controlled by both the transconductance ratios (W/L) and the clock frequency. The circuits are composed of analog filter banks whose impulse responses are the first derivative of Gaussian and its dilations. The transfer functions of the basic wavelet filter can be given by Pad approximation, which can decompose the transfer function into rational form so as to be conveniently implemented by SI circuits. Therefore, it can easily be applied to other filters with a prescribed impulse response. Simulated results indicated a good performance of wavelet transform.

High-temperature effects of AlGaN/GaN high-electron-mobility transistors on sapphire and semi-insulating SiC substrates

The high-electron-mobility transistors (HEMTs) have been demonstrated on both sapphire and semi-insulating (SI) SiC substrates, and the dc characteristics of the fabricated devices were examined at temperatures ranging from 25 to 500 °C. The decrease in drain current and the transconductance with the increase of temperature have been observed. The decrease ratio of transconductance and drain current was similar for both the HEMTs on sapphire and SI–SiC substrates at and above 300 °C. The HEMTs on SiC substrates showed better dc characteristics after being subjected to thermal stress up to 500 °C. Although the SiC-based HEMTs showed better characteristics up to the temperature of 300 °C, compared with the sapphire-based HEMTs, similar dc characteristics were observed on both at and above 300 °C. For high-temperature applications (⩾300 °C), additional cooling arrangements are essential for both devices.

Noise performance design of CMOS preamplifier for the active semiconductor neural probe

A systematic design guideline is presented for the noise performance of preamplifier for semiconductor neural probe which contains on-chip electronic circuitry. The overall signal-to-noise ratio (SNR) is calculated considering the spectral characteristics of the measured extracellular action potential and the low-frequency noise spectrum of the CMOS device from typical fabrication processes. An analytical expression of the output noise power is derived, and utilized to tailor the frequency response and device parameters which are controllable by the circuit designer. An analysis of the output SNR of a two-stage CMOS differential amplifier is given and the major factors which have significant effects on the SNR are determined. We showed that a little deviation of the input device sizes and transconductance ratio from the optimal values can significantly deteriorate the SNR. Quantitative information of the preamplifier circuit parameters for satisfactory noise performance is provided.

A Low-Power Wide-Linear-Range Transconductance Amplifier

The linear range of approximately ±75 mV of traditional subthreshold transconductance amplifiers is too small for certain applications—for example, for filters in electronic cochleas, where it is desirable to handle loud sounds without distortion and to have a large dynamic range. We describe a transconductance amplifier designed for low-power (< 1 µW) subthreshold operation with a wide input linear range. We obtain wide linear range by widening the tanh, or decreasing the ratio of transconductance to bias current, by a combination of four techniques. First, the well terminals of the input differential-pair transistors are used as the amplifier inputs. Then, feedback techniques known as source degeneration (a common technique) and gate degeneration (a new technique) provide further improvements. Finally, a novel bump-linearization technique extends the linear range even further. We present signal-flow diagrams for speedy analysis of such circuit techniques. Our transconductance reduction is achieved in a compact 13-transistor circuit without degrading other characteristics such as dc-input operating range. In a standard 2 µm process, we were able to obtain a linear range of ±1.7V. Using our wide-linear-range amplifier and a capacitor, we construct a follower–integrator with an experimental dynamic range of 65 dB. We show that, if the amplifier‘s noise is predominantly thermal, then an increase in its linear range increases the follower–integrator‘s dynamic range. If the amplifier‘s noise is predominantly 1/f, then an increase in its linear range has no effect on the follower–integrator‘s dynamic range. To preserve follower–integrator bandwidth, power consumption increases proportionately with an increase in the amplifier‘s linear range. We also present data for changes in the subthreshold exponential parameter with current level and with gate-to-bulk voltage that should be of interest to all low-power designers. We have described the use of our amplifier in a silicon cochlea [1, 2].

Design and fabrication of thin-film resistive SQUIDs

A partly resistive superconducting quantum interference device, or, RSQUID, is a device utilizing the nonstationary Josephson effect. The resistive part of the device makes it possible to apply a voltage to the Josephson junctions and, in this way, to convert it into frequency. The transconductance ratio of this conversion is defined by the physical constants: 1/o0 = 2e/h (o0 is the flux quantum). Five different types of thin-film RSQUIDs are described in this paper. They are: an rf RSQUID with an rf pump and one Josephson junction, a dc RSQUID with dc bias and two Josephson junctions, and three topologically different double RSQUIDs with an rf pump, two Josephson junctions and two quantization loops. All five types were successfully fabricated. The resistive part of all devices was an S-N-S sandwich with Nb as the superconductor (S) and Cu as the normal metal (N). Such a design of the resistive part sets its resistance on the order of a few tens of μΩ.

A SQUID with two quantization loops working in a negative detuning mode

The SQUID (superconducting quantum interference device) consists of two Josephson junctions incorporated into a superconducting loop. We attached another superconducting loop in parallel to the junctions. The magnetic flux to be measured is introduced into the first loop. The second one is coupled inductively to the tank circuit supplied by an rf current source. Previously we have reported that this configuration allows to pick up a signal voltage from the tank circuit up to 1 mV when the tank circuit is driven by its resonant frequency. In this paper we show that this kind of SQUID also allows to work at a negative detuning from the resonant frequency. It is found that the negative detuning mode improves the output signal voltage and transconductance ratio. The values of these parameters obtained experimentally are 1.8 mV and 9 mV/ϕ O, respectively (ϕ O is the flux quantum).

New method for assessment of depletion charge dependence of mobility in short-channel silicon MOS transistors

An original method for the extraction of the depletion charge mobility dependency coefficient α based on the exploitation of the body-to-gate transconductance ratio gb/gm(Vg) MOSFET characteristics is presented. This method allows the demonstration of the fact that α is a parameter strongly dependent on channel length for both device types (p or n). Moreover, it is shown that the channel length reduction of a is closely correlated to that of the depletion charge to the gate oxide capacitance ratio Cd/Cox due to the increase of charge sharing effect when scaling down the devices.

New folded cascode transconductor for bandpass ladder filters

The problems of designing CMOS transconductors for high frequency bandpass ladder filters are discussed, and an original circuit for overcoming these problems is presented. This circuit features a new grounded-quad input stage, a very high output impedance and the facility of low impedance inputs to avoid the requirement for transconductance ratios in scaled filters. The application of the new transconductor is illustrated by the design of a high frequency sixth-order Chebyshev bandpass filter. The performance of the transconductor and the filter are demonstrated by analysis and simulation results.

Generalized transconductance and transresistance methods for MOSFET characterization

The concept of transconductance is generalized to include substrate and drain transconductances in addition to the usual gate transconductance. Generalized transresistances are defined as the ratio of transconductance to square of conductance, and their proportionality to channel length and reciprocal square of gate drive is exploited to develop six new analytical methods for determination of MOSFET threshold voltage and channel length. Quantitative expressions for the error components due to short channel geometry effects, asymmetry and bias dependence of source and drain series resistances, and mobility degradation are derived and related to total threshold voltage and channel length errors of various methods. In one set of methods the channel length error due to bias dependence of series resistance is eliminated by extrapolation to zero gate drive. Estimated errors agree well with measurements on submicron LDD transistors and the high precision of the gate transresistance technique is demonstrated with a channel length error of 3 nm limited by instrument resolution. Threshold voltage accuracy of 2 mV is obtained with the square-root substrate transresistance technique when series resistance is independent of substrate bias.

Relation between incremental intrinsic capacitances and transconductances in MOS transistors

The ratio of intrinsic substrate and gate capacitances is related to the ratio of substrate transconductance to gate transconductance for a long-channel MOS transistor in strong inversion. It is shown that under certain assumptions the two ratios are approximately equal.

Neutron Degradation of Ion-Implanted and Uniformly Doped Enhancement Mode GaAs JFET's

The degradation of the saturation transconductance of ion-implanted GaAs enhancement mode JFETs by exposure to fast neutron fluences is calculated using simple analytical expressions published by bulk GaAs. It is shown that the saturation transconductance degrades linearly with neutron fluence at about the same rate for devices with widely varying implant profile and junction location, but equal gate geometry and similar threshold voltage. Thus the degradation rate of the relative transconductance (ratio of degraded to initial transconductances) with neutron fluence is decreased by a large initial transconductance. At a fixed gate voltage, and for a given threshold voltage and gate geometry, a large initial transconductance requires high peak dopant concentration and small mean deviation of the implant profile, and a gate junction located near the position of the peak dopant concentration. Experimental data for the degradation of the transconductance by neutron fluence for two ion-implanted transistors have been matched by theory assuming the presence of an undegraded parasitic source resistance. The degradation rate of an uniformly doped epitaxial transistor is found to be 79% of that of ionimplanted devices with equal gate geometry and similar threshold voltage.

A Wide-Band Inductive-Output Amplifier

In order to obtain amplification over a wide band of frequencies, a tube having high transconductance and low input and output capacitance is required. A grid-controlled inductive-output tube having these attributes and giving 10 watts output at 500 megacycles is described. A high ratio of transconductance to input capacitance is attained by close cathode-grid spacing and large grid-screen spacing, the latter made possible by use of a very high screen voltage. A low output capacitance is obtained by the use of an inductive-output arrangement which requires no power dissipation by the output electrodes. The design of the output circuit for low inherent capacitance is facilitated by the use of magnetic lenses to focus the electron beam. These lenses can be energized by a permanent magnet. Loading caused by secondary electron emission from the current collector is eliminated. The performance of the tube as a wide-band amplifier is described.