Input Noise Voltage Research Articles

Enhanced Sensitivity Cross-Correlation Method for Voltage Noise Measurements

Ultralow noise measurements often require the application of signal processing and correction techniques to go beyond the noise performances of front-end amplifiers. In this paper, a new method for the voltage noise measurement is proposed, which allows, at least in principle, the complete elimination of the noise introduced by the amplifiers used for the measurements. This is obtained by resorting to the conventional cross-correlation technique for the elimination of the contribution of the equivalent input voltage noise of the amplifiers and by using a new three-step-measurement procedure that exploits different amplifier-configuration measurements in order to subtract the contribution of the equivalent input current noise of the amplifiers

High-Tc dc SQUID readout electronics with low noise and high bandwidth

Using AD797 low noise op amps and 2SA1048 low noise transistors, we have developed a composite preamplifier for use in dc SQUID readout electronics. This preamplifier has a small dc drift and super low noise at high frequencies. The equivalent input voltage noise of the preamplifier is about 0.35nV/√Hz from 100kHz to 10MHz. Using this preamplifier, we developed dc SQUID readout electronics having low noise and high bandwidth. Used with a 3mm2 high-Tc dc SQUID, the white flux noise was about 18μΦ0/√Hz above 100kHz and the FLL bandwidth was about 2MHz. This readout electronics can be used for the applications of SQUID-based NDE and SQUID-based NQR.

A multifrequency magnetic induction tomography system using planar gradiometers: data collection and calibration

We developed a 14-channel multifrequency magnetic induction tomography system (MF-MIT) for biomedical applications. The excitation field is produced by a single coil and 14 planar gradiometers are used for signal detection. The object under measurement was rotated (16 steps per turn) to obtain a full data set for image reconstruction. We make measurements at frequencies from 50 kHz to 1 MHz using a single frequency excitation signal or a multifrequency signal containing several frequencies in this range. We used two acquisition boards giving a total of eight synchronous channels at a sample rate of 5 MS s−1 per channel. The real and imaginary parts of ΔB/B0 were calculated using coherent demodulation at all injected frequencies. Calibration, averaging and drift cancellation techniques were used before image reconstruction. A plastic tank filled with saline (D = 19 cm) and with conductive and/or paramagnetic perturbations was measured for calibration and test purposes. We used a FEM model and an eddy current solver to evaluate the experimental results and to reconstruct the images. Measured equivalent input noise voltage for each channel was 2 nV Hz−1/2. Using coherent demodulation, with an integration time of 20 ms, the measured STD for the magnitude was 7 nVrms (close to the theoretical value only taking into account the amplifier's thermal noise). For long acquisition times the drift in the signal produced a bigger effect than the input noise (typical STD was 10 nV with a maximum of 35 nV at one channel) but this effect was reduced using a drift cancellation technique based on averaging. We were able to image a 2 S m−1 agar sphere (D = 4 cm) inside the tank filled with saline of 1 S m−1.

64-channel magnetocardiogram system based on double relaxation oscillation SQUIDplanar gradiometers

We developed a magnetocardiogram (MCG) system measuring tangential field components usingplanar gradiometers. The SQUID insert has 64 planar gradiometers, of which 32 measure thex-component and theother 32 measure the y-component, where the chest surface is in thex and y plane. The SQUID sensor is a double relaxation oscillation SQUID type with highflux-to-voltage transfers. Thus, a room-temperature dc preamplifier with a modest inputvoltage noise could be used to detect the SQUID output voltage directly. The pickup coilsare first-order planar gradiometers with a baseline of 4 cm, and they were connected to theinput coils using a superconductive bonding of Nb wires. The sensor coverage area of the64 channels is 190 mm in diameter, and the average boil-off rate of liquid He is3.6 L/d in everyday operation of the SQUID system. The SQUID operation of the 64channels was done by digital controls and the average noise of the MCG system was4.6 fT Hz−1/2 at 100 Hz, inside a magnetically shielded room.

Measurement of Low-Frequency Noise of Modern Low-Noise Junction Field Effect Transistors

A description of the low-noise measurement system designed for experimental investigation of low-frequency noise of modern low-noise junction field effect transistors (JFETs) at a frequency range from 0.5 Hz to 50 kHz is presented. The noise floor of the system in terms of noise voltage spectral density is about 1.3 nV//spl radic/Hz at frequency 0.5 Hz, 0.6 nV//spl radic/Hz at frequency 1 Hz, 0.08 nV//spl radic/Hz at frequency 10 Hz, and /spl les/0.04 nV//spl radic/Hz at frequencies 100 Hz-50 kHz. Such a system makes it possible to measure 1/f noise and channel thermal noise of modern very low noise n-channel JFETs. Specifically, discrete samples of JFETs, IF9030, IF1801, IF3601, and 2N4338, were investigated at the frequency range mentioned above. The system provided negligible contribution of all noise sources except intrinsic noise of measured JFETs which was a 1/f noise at frequencies f < 10 Hz and a channel thermal noise at frequencies f /spl ges/ 10 Hz. The measured values of the equivalent input noise voltage spectral density e/sub n/ of the best samples from investigated JFETs IF9030, IF1801, IF3601, and 2N4338 were 2.7, 3.2, 3.4, and 22 nV//spl radic/Hz at frequency 0.5 Hz; 1.5, 2.1, 1.8, and 14 nV//spl radic/Hz at frequency 1 Hz; and 0.5, 0.4, 0.6, and 3.2 nV//spl radic/Hz at frequencies more than 1 kHz, correspondingly.

Noise Determination in Differential Pair-Based Second Generation Current Conveyors

The aim of this paper is to give simple design guidelines for the noise evaluation in differential pair-based second generation current conveyors (CCIIs). Input noise voltage and current have been expressed in terms of some well known constants and MOS gm values, so allowing a very fast estimation of the noise level. The presented theory has been verified, as an example, on two fabricated low voltage low power CCII topologies. Theoretical values, simulations and on chip measurements are also shown and their good agreement confirms the validity of the presented idea.

Integrated AQUID-gradiometer system for magneto-cardiography without magnetic shielding

First-order planar superconducting quantum interface device (SQUID) gradiometers were fabricated on the basis of the Nb/AlO/sub X//Nb technology developed at IPHT Jena. The gradiometers have two pickup loops each with quadratic dimensions of 2 cm and a baseline of 4 cm, integrated on chip. The intrinsic noise corresponds to a field resolution in one loop better than 3 fT//spl radic/Hz. Because of the high gradiometer balance (>10/sup 5/) the main signals masking a cardiogram are magnetic field gradients caused by nearby sources (laboratory equipment). To suppress the disturbances a second-order gradiometer was realized electronically. In addition, three washer-SQUID's arranged as a three-axis magnetometer were used to improve the disturbance suppression. All SQUID's were operated with directly coupled low-noise high-speed electronics controlled by a computer via a RS232 port. They have an input voltage noise of about 0.33 nV//spl radic/Hz, a 1/f corner frequency of 0.1 Hz and bandwidth of 5 MHz. The SQUID signals were digitized using a commercially available 16-bit ADC and digitally filtered. The system shows stable operation in the laboratory without magnetic shielding. A human cardiogram could be recorded in real time with a signal-to-noise ratio better than 20. First results of the measurement of foetal cardiogram in a magnetically shielded room are discussed.

A novel differential integrator for bioelectric events

A novel design of integrator with differential inputs suitable for bioelectric events is described. The main advantages of this integrator in comparison with the conventional differential integrators are that it uses a single RC network to control the integration action of the circuit, has higher quality factor and less equivalent input noise voltage. The obtained experimental and simulation results support the theoretical derivations.

Investigations on the low-power and low-frequency noise performance of pHEMT at 4.2 K

Pseudomorphic GaAs HEMTs with a gate length of 1μm have been realized and characterized at 4.2K. For the device with a gate width of 4mm, a gate leakage current less than 1pA and a total input capacitance lower than 10pF have been obtained for any practical bias conditions. The channel resistance can be modified more than 10 9 times with a gate bias variation less than a half volt. An intrinsic voltage gain higher than 10 can be reached with a power dissipation equal to or less than 0.15mW. Under this power supply condition, equivalent input noise voltages of 8.4, 3.2 and down to 1.3nV/√Hz can be obtained at frequencies of 1, 10 and 100kHz respectively. This noise voltage has been studied as a function of the drain current. Finally, the experimental results of this work have shown that the device's noise voltage can be linked to its DC parameters.

RF single electron transistor readout amplifiers for superconducting astronomical detectors of X-ray to sub-mm wavelengths

We have made Radio-Frequency Single-Electron Transistors (RF-SETs) with large input gates, and tested performance and modes of operation with the goal of using such devices as on-chip amplifiers for a variety of high impedance cryogenic photodetectors. We achieved /spl ap/100 kHz of closed-loop bandwidth for charge-locked-loop and transimpedance amplifier feedback configurations, and have combined amplifier outputs using a form of wavelength division multiplexing. With our choice of SET junction resistance, a 0.5 fF input gate capacitance gave a cotunneling-degraded charge noise of 1/spl times/10/sup -4/ e//spl radic/Hz, but a fairly low input voltage noise of 30 nV//spl radic/Hz.

Characterization of the Input Noise Sources of a dc SQUID

From the theory of the intrinsic noise in a dc SQUID by Tesche and Clarke, we derive the expressions of the current and voltage input noise spectral densities in a dc SQUID current amplifier operated in a flux locked mode. The expected current and voltage noises are compared, at audio frequencies, with the experimental results obtained with a low noise dc SQUID in which the input load (resonant and not) and the operating temperature (1–4 K) are changed. In order to evaluate the input voltage noise, which is directly related to the current noise around the SQUID loop and is usually neglected, we have used as the input circuit a LC resonator with a very high quality factor (≈106). Both the voltage and current input noises exceed the expected values by the same factor of about 8. This means that the modulus the optimum source impedance of the SQUID amplifier is still in agreement with the value expected from the theory, which is approximately given by the product of the input coil inductance and the angular frequency. To explain the excess noise results, we propose a model in which the voltage and current input noises are due to a thermal magnetic noise source which is present near the SQUID.

Correlation between equivalent input noise generators

Correlation between equivalent input voltage and current noise sources is usually considered to be unimportant by low-frequency circuit designers. This paper shows that measurements of noise figure as a function of source resistance make it possible to determine the spectral densities of the equivalent input noise generators and the correlation between them. Measurements were made on five different types of operational amplifier and considerable correlation was found in three of these.

A power-efficient, low-noise, wideband, integrated CMOS preamplifier for LSO/APD PET systems

A power-efficient, low-noise, wideband, 0.8-/spl mu/ CMOS charge-sensitive preamplifier has been designed and optimized for LSO/APD based PET systems. Low-noise, wideband performance is needed to achieve 1-ns FWHM timing resolution available from LSO/APD detectors, and power efficiency is needed since thousands of preamplifiers are required in a LSO/APD based PET tomograph. A methodology for optimizing wideband timing performance is presented to minimize dominant, frequency-turn-up, input-noise current associated with input noise voltage and total capacitance at the input. The methodology suggests the use of a nMOS input device operating in moderate inversion for high transconductance and low input white-noise voltage consistent with moderate input capacitance. Measured preamplifier input-referred, white-noise voltage is 0.65 mV//spl radic/Hz and risetime is 6 ns (58 MHz 3-dB bandwidth). Measured timing resolution is 1.15 ns FWHM and energy resolution is 12.1% FWHM for 511-keV gamma rays detected by a 4/spl times/4/spl times/10 mm LSO crystal coupled to an APD. Measurements include the APD capacitance of 30 pF and are obtained at a preamplifier input-device current of only 2 mA.

Low-noise electrometer and its low-noise cryogenic probe with completely guarded sample chamber

We describe a low-noise cryogenic probing system capable of measuring currents at the femtoAmpere (fA) level with integration time ⩽1 s and current noise spectra at the 1 fA/√Hz level with the sample cooled to cryogenic temperatures. The system consists of a low-noise electrometer and a low-noise cryogenic probe with completely guarded sample chamber. The low-noise electrometer has amplification of 0.1 V/1 pA with f3 dB=70 Hz, 1.1 V/pA with f3 dB=20 Hz, and/or 10.1 V/pA with f3 dB=1.4 Hz, and input current noise power density of 0.4 fA/√Hz rms and input voltage noise power density of ≈0.3 μV/√Hz rms at 1 Hz frequency. With the cryogenic probe, the entire measurement system has input current noise of 0.8 fA/√Hz rms at 1 Hz. The completely guarded sample chamber reduces parasitic conductance to &amp;lt;10−16 S and parasitic capacitance to ⩽15 fF between pairs of pins for current–voltage measurements. With this instrumentation, we are able to characterize the current and current noise of quantum well infrared photodetectors for sensitive zero-background applications.

A complete operational amplifier noise model: analysis and measurement of correlation coefficient

In contrast to the general operational amplifier (op amp) noise model widely used, we propose a more complete and applicable noise model, which considers the correlation between equivalent input voltage noise source e/sub n/ and current noise source i/sub n/. Based on the super-position theorem and equivalent circuit noise theory, our formulae for the equivalent input noise spectrum density of an op amp noise are applied to both the inverting and noninverting input terminals. By measurement, we demonstrate that the new expressions are significantly more accurate. In addition, details of the measurement method for our noise model parameters are given. A commercial operational amplifier (Burr-Brown OPA37A) is measured by means of a low-frequency noise power spectrum measuring system and the measured results of its noise model parameters, including the spectral correlation coefficient (SCC), are finally given.

A high CMRR current mode CMOS preamplifier for magnetic recording systems

This paper presents a high CMRR current mode CMOS preamplifier for a dual stripe magnetoresistive (DSMR) read head for magnetic recording systems. The current mode design approach a avoids high impedance nodes in the circuit and helps in achieving a high bandwidth. The differential circuit topology and the common mode negative feedback greatly improve the CMRR and PSRR. The preamplifier is designed using a 0.5 /spl mu/m CMOS technology. The achieved 3-dB bandwidth is 120 MHz at a gain of 45 dB, the CMRR and PSRR are 62 and 74 dB respectively. The equivalent input noise voltage is 1.83 nV//spl radic/Hz.

Electrode Impedance in Neurophysiologic Recordings: 1. Theory and Intrinsic Contributions to Noise

ABSTRACT.The theory behind common and differential mode signals, thermal noise, input voltage and input current noise, 1/f noise, and “popcorn” noise is discussed and the results applied to in vivo and in vitro electrophysiological recordings. Intrinsic noise in EEG and evoked potential paradigms was examined as a function of frequency, total electrode impedance, electrode impedance imbalance, and electrode type. It was demonstrated that electrode impedance may be much less important than other electrode properties in determining intrinsic noise levels. Dramatic increases in intrinsic noise levels below 1 Hz are demonstrated, both with real electrodes and pure resistances. Real electrodes, however, show much more rapid noise increase with decreasing frequency, probably as a result of complex electrochemical processes at the electrode-solution interface. From a practical perspective, although the noise levels in electrodes are much lower than the signal in human EEG, studies pursuing small changes in the EEG at low frequencies may benefit from careful attention to the type of electrode used for recording.

Compact low-voltage power-efficient operational amplifier cells for VLSI

Compact low-voltage power-efficient operational amplifiers are described that are very suitable as very-large-scale-integration library cells because of the small die area of 0.08 mm/sup 2/ and the minimum supply voltage of 1.8 V. A key part of the circuit is the rail-to-rail class-AB output stage with folded mesh feedback control that combines power efficiency with operation down to 1.8 V and allows sufficient gain in a compact two-stage topology. A version with rail-to-rail input stage features a rail-to-rail input range for supply voltages down to 2.5 V. The dc gain of the op amps is more than 80 db while driving 10 k/spl Omega/, and the unity-gain frequency is 4 MHz with phase margin of 67/spl deg/ while driving 5 pF. The equivalent input noise voltage is 38 nV//spl radic/(Hz) at a frequency of 100 kHz. The amplifiers have been implemented in a standard digital 1.6-/spl mu/m complementary metal-oxide-semiconductor process.

A CMOS channel-select filter for a direct-conversion wireless receiver

A highly selective and linear switched-capacitor channel-select filter is fabricated in 1-/spl mu/m CMOS for a direct-conversion wireless receiver operating in the 902-928 MHz ISM band. The filter selects a 230-kHz wide channel and attenuates by at least 50 dB from 320 kHz to 57 MHz. The input IP3 is +30 dBm, the input-referred noise in the passband is 70 nV//spl radic/Hz, and the circuit takes 4.6 mA from a 3.3 V supply. Direct subsampling of the 915 MHz RF input signal by the filter front-end is also demonstrated with only a small degradation in linearity. The input noise voltage is halved in a redesign while keeping the current drain unchanged.

A 1-GHz bipolar class-AB operational amplifier with multipath nested Miller compensation for 76-dB gain

A 1-GHz operational amplifier with a gain of 76 dB while driving a 50-/spl Omega/ load is presented. The equivalent input noise voltage is as low as 1.2 nV//spl radic/Hz. This combination of extremely high bandwidth, high gain, and low noise is the result of a three-stage all-n-p-n topology combined with a multipath nested Miller compensation. Using 10-GHz f/sub T/ n-p-n transistors, the realizable bandwidth could be of the order of 2-3 GHz. However, bond-wire inductances restrict the useful bandwidth to 1 GHz. The amplifier occupies an active area of 0.26 mm/sup 2/ and has been realized in the bipolar part of a 1-/spl mu/m BiCMOS process.