Excess Noise Sources Research Articles

Excess Lorentzian Noise in Partially Depleted SOI nMOSFETs Induced by an Accumulation Back-Gate Bias

It is shown that the application of an accumulation back-gate bias (V/sub GB/) to partially depleted silicon-on-insulator nMOSFETs induces an excess Lorentzian noise component in the front-channel current noise spectral density. The effect is observed for front-gate biases (V/sub GF/) from weak to strong inversion. Below the threshold for electron valence band (EVB) tunnelling, the parameters of the Lorentzian noise are independent on V/sub GF/, but strong functions of V/sub GB/. Given the similarity with the behaviour of the EVB tunnelling-related noise overshoot, the observations are interpreted in a similar way. It is proposed that the additional source of floating-body related excess Lorentzian noise is shot and thermal noise in the body charge, filtered by the RC impedance of the source-body junction. The possible origin of the body-charge fluctuations and related shot and thermal noise will be discussed.

Critical Discussion of the Front–Back Gate Coupling Effect on the Low-Frequency Noise in Fully Depleted SOI MOSFETs

The low-frequency (LF) noise in fully depleted (FD) silicon-on-insulator (SOI) MOSFETs is investigated both theoretically and experimentally, with special emphasis on the front-back gate coupling effect. The impact of different technological parameters on the 1/f noise coupling factor has been investigated, within the frame of a number fluctuations model, ascribing the 1/f noise to traps in the front and buried oxide. It is shown that the coupling effect significantly increases for a smaller film thickness, while the influence of the buried-oxide thickness is less pronounced. The resulting 1/f noise is strongly determined by the ratio of the back and front oxide trap densities and is expected to be higher for the back interface in depletion compared with accumulation. These trends have been verified for FD SOI MOSFETs fabricated in a 0.1-/spl mu/m CMOS technology and measured in linear operation. Except for long-channel transistors and for a front-gate bias <1 V in absolute value, a higher noise is usually found for the back-gate bias in accumulation. It is shown that in the latter case, additional noise sources become active, which drastically enhance the LF noise. Finally, the origin and back-gate bias dependence of these excess noise sources will be critically discussed.

On the nature of the superconducting-to-normal transition in transition edge sensors

A transition edge sensor (TES) is an X-ray microcalorimeter based on a thin film superconductor voltage-biased within its resistive transition. This superconductor is usually modelled as a variable resistor, completely described by a single value of the logarithmic temperature coefficient, α. Recent measurements of excess noise (and poorer than expected X-ray energy resolution) demand, however, a more detailed physical model of the superconducting-to-normal transition. In this paper, we show, using data from Ir films and from Ti- , W- , Ir- , Mo/Cu- and Ti/Au-based devices, that the TES may be plausibly regarded as a two-dimensional superconductor exhibiting a Kosterlitz–Thouless–Berezinsky phase transition due to vortex dynamics. Adopting this model leads to : (i) expressions for the variation of both TES resistance and the α-parameter with temperature (or bias voltage) within the transition and (ii) a physical description of the excess noise source and its scaling laws, together with strategies for its reduction.

Influence of magnetic field on 1/ f noise and thermal noise in multi-terminal homogeneous semiconductor resistors and discrimination between the number fluctuation model and the mobility fluctuation model for 1/ f noise in bulk semiconductors

We have derived an accurate noise current density equation for homogeneous semiconductors under a constant magnetic field with three noise source terms which are the thermal noise source term and the two excess noise source terms due to either carrier number fluctuation or carrier mobility fluctuation. Based on Hooge’s empirical relations for 1/ f noise, the two excess noise source terms are shown to become equal to each other under zero magnetic field. Using the characteristic potential method and employing Hooge’s empirical relations for 1/ f noise, we have derived the formulas for the short-circuit terminal 1/ f noise and thermal noise currents and the open-circuit terminal 1/ f noise and thermal noise voltages of multi-terminal homogeneous semiconductor resistors with arbitrarily-shaped 2-D geometries under a constant magnetic field. We have shown that the derived formulas can explain the measured 1/ f noise and thermal noise of n-GaAs rectangularly-shaped Hall devices and n-GaAs Corbino disks under magnetic field from 0 to 8 T at room temperature. It is also shown that the magnetic field dependence of 1/ f noise in bulk semiconductors should be explained by the number fluctuation model rather than by the mobility fluctuation model.

Comparison of low-frequency noise in III–V and Si∕SiGe HBTs

The low-frequency noise characteristics of double self-aligned InP/InGaAs and two types of Si/SiGe heterojunction bipolar transistors (HBTs) were investigated. Spectral analysis shows no striking differences; the spectra are composed of a 1/f component and the white noise is always reached at low biases. A general trend for all the transistors was the presence of Lorentzian component(s) for the smallest devices. The voltage coherence function was always unity for SiGe transistors; and for the first time, it was found to be close to zero for InP devices. Concerning the 1/f noise level, both types of transistors have approximately a quadratic dependence on base current bias and an inverse dependence on the emitter area. Thus, a comparison of the 1/f noise level has been made using the K/sub b/ parameter, and values around 10/sup -9/ /spl mu/m/sup 2/ for SiGe HBTs and around 10/sup -8/ /spl mu/m/sup 2/ for InP HBTs were found. These results are of the same order of magnitude as the best published ones. The low-frequency noise results suggest that excess noise sources are mainly located at the intrinsic emitter-base junction for the two types of SiGe devices, and, for the InP HBTs, a correlated noise source is located at the emitter periphery. To compare different devices and technologies, f/sub c//f/sub T/ was studied as a function of collector current density and for some HBT technologies f/sub c//f/sub T//spl prop/J/sub c/ (f/sub c/ is corner frequency at which the white noise and 1/f noise are equal and f/sub T/ is the unity current gain frequency). The effects of different processing conditions, designs and temperature were also investigated and discussed.

Transition edge X-ray sensors for industrial applications

We present the performance of transition edge sensors (TES) using Au/Ti bilayer (operating temperature 113 mK ) with fast decay time for industrial applications. The energy resolution was 9.2 eV for Mn K α1 (designed value 3.5 eV ) and the decay time 71 μs , yielding the count rate 2.3 kcps (count per sec). We found that the energy resolution was limited by excess noise. The excess noise increased in proportion to the inverse of TES resistance as decreasing the TES resistance from normal resistance. The source of excess noise might exist in the TES.

Eliminating the vibrational noise in continuously filled 1 K pots

We present a study on the origin of the vibrational noise originating from pumped helium chambers (1 K pots) that are continuously replenished from a main bath at 4.2 K. The vibrations can be eliminated by thermalizing the helium, coming from the main bath, to the pot temperature. Vibrations in 1 K pots are a source of excess noise in cryogenic detectors and therefore have detrimental consequences on their performance.

Physical noise modeling of SOI MOSFETs with analysis of the Lorentzian component in the low-frequency noise spectrum

The implementation of a general physics-based compact model for noise in silicon-on-insulator (SOI) MOSFETs is described. Good agreement is shown between model-predicted and measured low-frequency (LF) noise spectra. In particular, the behavior of an excess Lorentzian component that dominates the LF noise spectra of SOI MOSFETs is investigated. Shot noise associated with the generation and removal (via recombination or a body contact) of body charge is shown to underlie the behavior of the Lorentzian in both floating-body and body-tied-to-source SOI MOSFET's operating under partially depleted or mildly fully depleted conditions; the Lorentzian is suppressed when the body is strongly fully depleted. Good physical insight distinguishes the behavior of the Lorentzian components in all these devices, and predicts the occurrence of additional excess noise sources in future scaled technologies. Simple analytic expressions that approximate the full model are derived to provide the insight.

Full observation of single-atom dynamics in cavity QED

We report the use of broadband heterodyne spectroscopy to perform continuous measurement of the interaction energy between one atom and a high-finesse optical cavity, during individual transit events of $\sim 250$ $\mu$s duration. Measurements over a wide range of atom-cavity detunings reveal the transition from resonant to dispersive coupling, via the transfer of atom-induced signals from the amplitude to the phase of light transmitted through the cavity. By suppressing all sources of excess technical noise, we approach a measurement regime in which the broadband photocurrent may be interpreted as a classical record of conditional quantum evolution in the sense of recently developed quantum trajectory theories.

Noise-free quantum-nondemolition measurement using optical solitons

The collision of solitons in optical fibers is a promising scheme for the quantum-nondemolition measurement of the photon number using the cross Kerr effect. In a realistic experimental situation, the sensitivity of the measurement is limited not only by self-phase-modulation noise of the probe but also by the phase noise added due to the vibrational excitations of the fiber. Former schemes proposed to overcome either of these problems are not compatible with each other. In this paper the sensitivity of different type of detection is carefully studied. We propose a solution based on the use of a phase-shifting cavity. We show that this provides a conceptually and practically simple solution eliminating simultaneously both sources of excess noise.

Practical aspects for SQUID applications

Classical applications of superconducting quantum interference devices (SQUIDs), like biomagnetic measurements or precision voltage and current measurements have in almost all cases been carried out in a magnetically and electrically well shielded environment. Recently, the demonstration of highly sensitive HTS SQUIDs has increased the interest in SQUIDs for a broader range of applications like geophysical measurements or nondestructive materials evaluation. The operation of SQUIDs in an unshielded environment, however, leads in many cases to a significant reduction in their sensitivity due to numerous sources of interference and excess noise. The paper reviews recent work on aspects related to the operation of SQUIDs in unshielded environments.

Wideband on-wafer noise measurement setup for noise characterization of active devices in the low VHF band

This paper addresses a new wideband on-wafer measurement test set designed for noise characterization of microwave active devices over the frequency range of 300 kHz to 150 MHz. Noise parameters obtained from the multiple impedance technique on a GaAlAs/GaAs heterojunction bipolar transistor (HBT) from 300 kHz to 70 MHz are reported and compared with low-frequency noise data. Investigation of the excess noise sources of III-V HBT's is performed well above the 100 kHz frequency limit of standard dynamic signal analyzers and noise modeling of these devices is reported.

Signal resolution of RSFQ comparators

We have designed and tested a Rapid Single-Flux-Quantum (RSFQ) circuit for experimental measurement of the fluctuation-induced threshold uncertainty of switching of SFQ comparators, which are essential components of RSFQ logic devices, analog-to-digital converters, and digital SQUIDs. Statistical density of the switching events in comparators using externally-overdamped Nb-trilayer Josephson junctions has been found to be in agreement with the distribution which follows from a simple theory based on a time-dependent harmonic-oscillator model of the device. Width of the distribution (i.e, the single-shot current resolution of the comparator) measured as a function of temperature in the range 1.7-4.2 K corresponds to fundamental (thermal/quantum) fluctuations, with no evidence of excess noise sources.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Electron-electron scattering in quantum wires and its possible suppression due to spin effects.

A microscopic picture of electron-electron pair scattering in single mode quantum wires is introduced which includes electron spin. A new source of excess noise for hot carriers is presented. We show that zero magnetic field spin splitting in quantum wires can lead to a dramatic spin-subband dependence of electron-electron scattering, including the possibility of strong suppression. As a consequence extremely long electron coherence lengths and new spin-related phenomena are predicted. Since electron bands in III-V semiconductor quantum wires are in general spin split in zero applied magnetic field, these new transport effects are of general importance.

Elimination of phase noise of pulsed laser diodes in electrooptic sampling systems

The multimode operation of pulsed laser diode sources is converted into excess noise in electrooptic probing systems for electrical waveforms and prevents measurements with shot-noise-limited sensitivity. The transfer function of the sampling system is determined and analyzed in terms of the wavelength and electric field dependence. It is shown that by proper orientation of the polarizing components within the optical beam path, points of operation can by found which have both a vanishing wavelength dependence of the transfer function and maximum sensitivity to the electric fields to be measured. Excess noise sources due to wavelength fluctuations of the laser are thus eliminated, and shot-noise-limited voltage resolution is obtained. The system has been optimized to yield a voltage sensitivity of 2.3 mV/ square root Hz for a longitudinal LiNbO/sub 3/ probe crystal. >

The loss of signal to noise due to imperfect isolation between the channels of a quadrature nuclear magnetic resonance probe

We consider a quadrature nuclear magnetic resonance (NMR) probe with imperfect isolation between the two channels arising from a parasitic mutual impedance, and show that the theoretically achievable signal-to-noise ratio is reduced by a factor of (1+η2)−1/2, relative to that of a probe with perfect isolation, where −20 log η is the isolation in dB; provided that all relevant receiver components are matched to a common impedance. The theoretical result is verified by experimental determination of the response of actual NMR imaging probes when driven with coherent signal, and the response of model circuits simulating actual probes when driven with sources of excess noise; the simulated circuits are needed to reduce extraneous interference in the noise measurements. The theory is also confirmed by computer simulations of noise response in coupled resonator circuits.

Mode Switching in Injection Lasers Induced by Temperature Variation and Optical Feedback

Effects of optical feedback and temperature fluctuation on the longitudinal mode behavior of an AlGaAs injection laser have been studied utilizing the transient temperature variation produced by pulsed current excitation. Small temperature fluctuations (∼0.01 K for a feedback distance of about 300 mm) induce mode switching at relatively low power levels (1∼4 mW). This behavior is shown to be consistent with the compound cavity model and the gain suppression effect associated with the inhomogeneity of the gain spectrum. It has also been found that mode switching is always accompanied by a change in the output power, indicating that this phenomenon is an important source of excess optical noise.

Material from the Internal Surface of Squid Axon Exhibits Excess Noise: Implications in Modeling Membrane Noise

A fluid material from a squid (Loligo pealei) axon was isolated by mechanical application of two types of microcapillary (1-3-mum Diam) to the internal surface of intact and cut-axon preparations. Current noise in the isolated material exceeded thermal levels and power spectra were 1/f in form in the frequency range 1.25-500 Hz with voltage-dependent intensities that were unrelated to specific ion channels. Whether conduction in this material is a significant source of excess noise during axon conduction remains to be determined. Nevertheless, a source of excess noise external to or within an ion channel may not be properly represented solely as an additive term to the spectrum of ion channel noise; a deconvolution of these spectral components may be required for modeling purposes.

Excess noise sources due to defects in forward biased junctions : G. Blasquez. Solid-St. Electron.21, 1425 (1978)

Excess noise sources due to defects in forward biased junctions : G. Blasquez. Solid-St. Electron.21, 1425 (1978)

A new look at an old idea on 1/f noise in semiconductors

Two-electron centres are suggested as plausible candidates for a source of excess current noise in semiconductors. The noise level estimated under the assumption that it is produced by modulation of carrier concentration, appears to be slightly too small.