Abstract

This paper presents a 16-bit 1 MS/s pseudo-differential Successive-Approximation-Register Analog-to-Digital Converter (SAR ADC) achieving an ENOB of 15-bit. To accommodate the pseudodifferential input, a differential DAC utilizing both monotonic and traditional switching is designed. For both dynamic and static performance improvement, three techniques are proposed. First, a foreground digital selfcalibration method is described to eliminate the capacitor mismatch errors. Some of the LSBs capacitors are utilized to measure and calculate the bit weights of other capacitors. Second, a DNL enhancement technique is presented. The fractional value capacitors are used to subtract an analog voltage from the DAC before the conversion is finished, which cancels out the consequent effect when the fractional part of the digital output is discarded before final output. Third, a low-offset low-noise comparator is designed. A reset timer with DAC settling replica is proposed to make the pre-amplifier in the comparator to start to amplify the DAC summing node voltage right after the DAC has fully settled. A prototype ADC is fabricated in a 0.18-μm 5-V CMOS process. It measures a 92.3-dB SNDR and a 107.9-dB SFDR. The DNL and INL are within ±0.3 LSB and ±0.72 LSB, respectively. The overall power consumption, drawn from the 5 V power supply, is 40 mW.

Highlights

  • High-resolution successive-approximation-register analogto-digital converters (SAR ADC) with several hundreds of kS/s or several MS/s are extensively used in industrial measurement, medical instruments, and battery-powered systems due to its simple structure and low power

  • A 16-bit 1 MS/s pseudo-differential SAR ADC is presented in this paper

  • The bit weights errors are fixed by a digital self-calibration technique which utilizes some of the LSBs capacitors to measure and calculate the bit weights of other capacitors

Read more

Summary

INTRODUCTION

High-resolution successive-approximation-register analogto-digital converters (SAR ADC) with several hundreds of kS/s or several MS/s are extensively used in industrial measurement, medical instruments, and battery-powered systems due to its simple structure and low power. A 16-bit pseudo-differential SAR ADC with redundancy is introduced. The SAR logic in calibration, including the redundant bits, is designed to be identical with that in normal conversion, so as to simplify the logic circuits. Based on the digital bit weight calibration, a DNL enhancement technique is proposed in this paper to further improve the ADC performance by utilizing the fractional part of the accumulated output codes. Another major issue to be considered is that the foreground digital calibration requires the comparator offset to be eliminated. The clock rising edge immediately after the sampling point at the bottom plate of CDMY raises the DAC voltage to prevent it from dropping below zero

SWITCHING METHOD
NOISE CONSIDERATION
DNL ENHANCEMENT
COMPARATOR DESIGN
PROTOTYPE DESIGN AND MEASUREMENT RESULTS
Findings
CONCLUSION
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call