Switched Biasing" is proposed as a new circuit technique that exploits an intriguing physical efect: cycling a MOS transistor between strong inversion and accumulation reduces its intrinsic 14 noise. The technique is implemented in a 0 . 8 p CMOS sawtooth oscillator by periodically of-switching of the bias currents during time intervals that theyare not contributing to the circuit operation. Measurements show a reduction of the lynoise induced phase noise by more than 8 dB, while the power consumption is reduced by more than 30% as well.

This paper gives experimental proof of an intriguing physical effect: periodic on-off switching of MOS transistors in a CMOS ring oscillator reduces their intrinsic 1=f noise and hence the oscillator's close-in phase noise. More specifically, it is shown that the 1=f 3 phase noise is dependent on the gate-source voltage of the MOS transistors in the off state. Measurement results, corrected for waveform-dependent upconversion and effective bias, show an 8-dB-lower 1=f 3 phase noise than expected. It will be shown that this can be attributed to the intrinsic 1=f noise reduction effect due to periodic on-off switching.

No abstract

In this paper, we present measurements and simulation of random telegraph signal (RTS) noise in n-channel MOS-FETs under periodic large signal gate-source excitation (switched bias conditions). This is particularly relevant to analog CMOS circuit design where large signal swings occur and where LF noise is often a limiting factor in the performance of the circuit. Measurements show that, compared to steady-state bias conditions, RTS noise can decrease but also increase when the device is subjected to switched bias conditions. We show that the simple model of a stationary noise generating process whose output is modulated by the bias voltage is not sufficient to explain the switched bias measurement results. Rather, we propose a model based on cyclostationary RTS noise generation. Using our model, we can correctly model a variety of different types of LF noise behavior that different MOS-FETs exhibit under switched bias conditions. We show that the measurement results can be explained using realistic values for the bias dependency of and .

In this work, we study random telegraph signal ͑RTS͒ noise in metal-oxide-semiconductor field effect transistors when the device is periodically and rapidly cycled between an "on" and an "off" bias state. We derive the effective RTS time constants for this case using Shockley-Read-Hall statistics applied under transient conditions. In this way, we show that the oft-observed reduction in RTS noise under such bias conditions can be explained by a nonuniform ͑e.g., U-shaped͒ distribution in energy of interface traps.

Abstract-In small-area MOSFETs widely used in analog and RF circuit design, low-frequency (LF) noise behavior is increasingly dominated by single-electron effects. In this paper, we review the limitations of current compact noise models which do not model such single-electron effects. We present measurement results that illustrate typical LF noise behavior in small-area MOSFETs, and a model based on Shockley-Read-Hall statistics to explain the behavior. Finally, we treat practical examples that illustrate the relevance of these effects to analog circuit design. To the analog circuit designer, awareness of these single-electron noise phenomena is crucial if optimal circuits are to be designed, especially since the effects can aid in low-noise circuit design if used properly, while they may be detrimental to performance if inadvertently applied.Index Terms-CMOS, flicker noise, large-signal excitation, lowfrequency noise, low-noise circuit design, MOSFET, noise reduction, RTS noise, switched biasing, 1 noise.

No abstract

The behavior of RTS noise in MOSFETs under large-signal excitation is experimentally studied. Our measurements show a significant transient effect, in line with earlier reports. We present a new physical model to describe this transient behavior and to predict RTS noise in MOSFETs under large signal excitation. With only three model parameters the behavior is well described, contrary to existing models.