Six-bit 2.7-GS/s 5.4-mW Nyquist complementary metal-oxide semiconductor digital-to-analogue converter for ultra-wideband transceivers

Abstract

This study presents a 6-bit 2.7 GS/s low-power digital-to-analogue converter (DAC) for ultra-wideband transceivers. A '2(thermometer) + 4(binary)' segmented architecture is chosen to reach a compromise between the current source cell's area and the operating speed of the thermometer decoder. In addition, the proposed pseudo-thermometer structure improves the DAC's dynamic performance. The bipolar current source cell and latch clock delay technique are employed to reduce the power consumption in the analogue and digital parts, respectively. Moreover, the compact de-glitch latch presented in this study simplifies the conventional latch design and layout. This DAC was implemented in a standard 0.13 mm 1P8M complementary metal-oxide semiconductor technology with the active area of 0.0585 mm 2. The measured differential non-linearity and integral non-linearity are less than 0.09 and 0.11 least significant bit, respectively. The measured spurious-free dynamic range is more than 36 dB over the Nyquist frequency at the sampling frequency of 2.7 GHz. The DAC consumes 5.4 mW with a near-Nyquist sinusoidal output at 2.7 GS/s, resulting in a better figure of merit of 31 fJ/conversion-step than other published arts.

A 6-bit 4 GS/s pseudo-thermometer segmented CMOS DAC

“…Compared to the pseudo-segmentation decoding method, a better solution called pseudo-thermometer decoding method in Ref. [3] is chosen to tackle the large current source area issue that exists in the pseudo-segmentation decoder. The high conversion rate, low power consumption and the relatively low f T based on the standard 0.18 m CMOS technology all make the design a challenge.…”

A 6-Bit 5GS/s Pseudo-Thermometer Segmented CMOS DAC

“…In [1,2] a pseudo-segmentation decoding method is proposed to improve dynamic performance instead of a traditional thermometer decoding. A better solution called pseudo-thermometer decoding method in [3] is chosen to tackle the large current source area issue exists in the pseudo-segmentation decoder.…”