Search

In this article, a transient-enhanced fully synthesizable digital low dropout regulator (FS-DLDO) is proposed for ultra-low-power applications. The FS-DLDO uses a fully synthesizable comparator (FS-Com) to sense load variations. A digital logic controller (D-CTRL) tunes the output voltage (VO) through a quad-loop architecture. The quad-loop architecture uses short bidirectional shift registers (BSR) to achieve fast-transient response and reduce leakage current. In addition, the FS-DLDO supports freeze-mode to regulate a ripple-free VO and minimize power consumption at a steady state. To demonstrate this entire design using standard-cells, the P-MOSFET array (PTA) used in traditional digital low dropout regulator (DLDO) is replaced with an array of three-state buffers (TSA). The layout is created using digital design flow in TSMC CMOS 45 nm process, which occupies a 6708 µm2 area. For a power supply (VSUP) range of 0.5–1 V, the FS-DLDO can provide regulated VO with a 50 mV dropout voltage. At VSUP = 500 mV and clock frequency (fCLK) of 10 MHz, the proposed regulator achieves a transient response time of 0.91 µs. This prototype achieves a peak current efficiency of 99.90% and produces a ripple-free VO at a steady state. [ABSTRACT FROM AUTHOR]

Electrocardiogram (ECG) signal processing and analysis is becoming more and more popular as it is useful in diagnosis and prognosis of human heart and clinically automatic machine estimation is based upon it. R-peak is the most important component in ECG beat and is widely used to investigate normal and abnormal subjects (patients). From the last few decades, R-peak detection in ECG has been the most challenging topic in the biomedical research. As QRS complex has high frequency in ECG as compared to other waves (P, T, U-wave), so majority of algorithms estimate QRS complex by either filtering or suppressing the lower frequency waves, including various artifacts like baseline wander, power line interference, and electromyograph noises. This paper demonstrates a new kind of ECG denoising algorithm based on self-convolution window (SCW) concept. The SCW based on Hamming window, herein referred to as Hamming self-convolution window, is used to design a new kind of filter which possesses negligible ripples in the stop band, as compared to the conventional window-based filters. This algorithm is validated on MIT-BIH arrhythmia database and the results outperform in terms of sensitivity, positive predictivity, and error rate obtained as 99.93%, 99.95%, and 0.117%, respectively, as compared to the other well-established works. The approach has also outperformed the results of well-established window-based filters (Hamming and Kaiser) in terms of reduced false negative, false positive, and error rate. [ABSTRACT FROM AUTHOR]