In this paper, a 402 MHz reconfigurable regulating rectifier (R3) with a wide input dynamic range has been designed for wireless power transfer (WPT) systems. The rectifier has a 6-stage cross-coupled (CC) rectifier driven by a differential RF input. With the assistance of the analog regulation circuit, the rectifier achieves a regulated 1.8V DC output voltage with a ripple factor of less than 0.07%. Moreover, a digital control logic is employed to modify the stages of the R3to balance the voltage conversion ratio (VCR) and power conversion efficiency (PCE) according to the input signal. The proposed R3can achieve a VCR of 3.66 at a load of 500 kΩ and a PCE of 60.1% at -11.3 dBm output power.

Bootstrapped switches are extensively used in applications demanding high-speed and high-linearity sampling. Nevertheless, the mathematical analysis of distortions caused by parasitic effects has not been performed previously. This paper essentially analyzes three factors that affect the linearity of the switch: source-drain exchange, parasitic capacitive division, and charge redistribution. By employing the time-domain waveform integration technique along with Chebyshev polynomials, the Fourier series expression for on-resistance of the switch can be effectively elucidated. Consequently, the effects of all sources of distortion have been quantified individually. It has been found that both the on-resistance and the bootstrap capacitor has a quadratic effect (40 dB/dec) on the third-harmonic distortion of the sampled signals. A$1\times 9$bootstrapped switch array has been fabricated using 180 nm technology to verify the proposed calculation model. Using the subsampling method, the measured total harmonic distortion (THD) performance was found to be highly consistent with the model calculation results, demonstrating the effectiveness and accuracy of the calculation model. The proposed calculation model will be highly beneficial for designers of bootstrapped switches.

Recent developments in large language models (LLM) and generative AI have unleashed the astonishing capabilities of text-to-image generation systems to synthesize high-quality images that are faithful to a given reference text, known as a “prompt”. These systems have immediately received lots of attention from researchers, creators, and common users. Despite the plenty of efforts to improve the generative models, there is limited work on understanding the information needs of the users of these systems at scale. We conduct the first comprehensive analysis of large-scale prompt logs collected from multiple text-to-image generation systems. Our work is analogous to analyzing the query logs of Web search engines, a line of work that has made critical contributions to the glory of the Web search industry and research. Compared with Web search queries, text-to-image prompts are significantly longer, often organized into special structures that consist of the subject, form, and intent of the generation tasks and present unique categories of information needs. Users make more edits within creation sessions, which present remarkable exploratory patterns. There is also a considerable gap between the user-input prompts and the captions of the images included in the open training data of the generative models. Our findings provide concrete implications on how to improve text-to-image generation systems for creation purposes.

In this paper we propose a novel high-resolution chemical sensing front-end with integrated sigma delta quantisation, and the chemical sensing capability is realised by the intrinsic passivation layer. A charge transfer amplifier with adjustable pre-amplification has been adopted to not only transmit the chemical signal to the next sampling stage, but also alleviate the kT/C thermal noise requirement. A pseudo-resistor is added to provide the DC operating points for the front-end, and it is implemented via a MOS transistor applying direct tunnelling effect to achieve huge resistance value. Simulation results show that through oversampling and noise shaping functions, the final resolution can reach up to 90.1 dB even with a 0.3V input signal, making it suitable for applications requiring high-precision chemical detection.