Analog Devices International Unlimited Company

Analog Devices Global

Analog Devices,inc.

Jialin Zhao

Hajime Shibata

Yunzhi Dong

Wenhua William Yang

Jose Barreiro Silva

Richard E Schreier

Gil Engel

Trevor Clifford Caldwell

Zhao Li

Hao Luo

David Nelson Alldred

Victor Kozlov

Steven C Rose

Prawal Man Shrestha

Qingdong Meng

Lin Zhang

Donald W Paterson

Shawn S Kuo

Martin McCormick

Gerard E Taylor

Yuhu Chen

Tommi Wang

Dong Li

Digital to analog converter generates an analog output corresponding to a digital input by controlling DAC cells using bits of the digital input. The DAC cells individually make a contribution to the analog output. Due to process, voltage, and temperature variations, the DAC cells may have duty cycle error or mismatches. To compensate for the duty cycle error of a DAC cell, a small amount of charge is injected into a low-impedance node of a DAC cell when the data signal driving the DAC cell transitions, or changes state. The small amount of charge is generated using a capacitive T-network, and the polarity of the charge injected is opposite of the error charge caused by duty cycle error. The opposite amount of charge thus compensates or cancels out the duty cycle error, and duty cycle error present at the output of the DAC cell is reduced.

DAC duty cycle error correction

Digital to analog conversion generates an analog output corresponding to a digital input by controlling unit elements or cells using data bits of the digital input. The unit elements or cells individually make a contribution to the analog output. Due to process, voltage, and temperature variations, the unit elements or cells may have mismatches and/or errors. The mismatches and/or errors can degrade the quality of the analog output. To extract the mismatches and/or errors, a transparent dither can be used. The mismatches and/or errors can be extracted by observing the analog output, and performing a cross-correlation of the observed output with a switching bit stream of the dither. Once extracted, the unit elements or cells can be adjusted accordingly to reduce the respective mismatches and/or errors.

Background duty cycle error measurement for RF DAC

Digital to analog conversion generates an analog output corresponding to a digital input by controlling unit elements or cells using data bits of the digital input. The unit elements or cells individually make a contribution to the analog output. Due to process, voltage, and temperature variations, the unit elements or cells may have mismatches. The mismatches can degrade the quality of the analog output. To extract the mismatches, a transparent dither can be used. The mismatches can be extracted by observing the analog output, and performing a cross-correlation of the observed output with the dither. Once extracted, the unit elements or cells can be adjusted accordingly to reduce the mismatches.

Background timing skew error measurement for RF DAC

Background static error measurement and timing skew error measurement for RF DAC

A DAC has a plurality DAC cells, and timing mismatch among the DAC cells can introduce errors in an output of a DAC. An efficient technique can be implemented to extract the timing error of a DAC cell. The technique involves a mixer to mix the timing error to DC (DC stands for direct current, where signal frequency is zero) and a VCO ADC to observe the output of the DAC cell to measure and extract the timing error. A first measurement is made using a first quadrature phase signal and a second measurement is made using a second quadrature phase signal. A difference between the first measurement and the second measurement yields the timing error of the DAC cell. Advantageously, the mixer can be integrated within a voltage-to-current converter of the VCO ADC. The timing error can be corrected in the digital domain or analog domain.

DAC calibration using VCO ADC

Adaptive toggle number compensation techniques for reducing data dependent supply noise in DACs are disclosed. Various embodiments are based on setting a certain target toggle number for a plurality of DAC units used to convert at least a portion of a digital data sample and then applying various adaptive techniques to try to achieve the target toggle number in converting the data sample from digital to analog domain. Adaptive toggle number compensation techniques described herein try to reduce data dependent supply noise by deliberately limiting, to a certain target number, the number of DAC units that undergo a switch from the digital input of 1 to 0 or from 0 to 1 in converting a digital data sample. Compared to the conventional dummy signal generation approach, such adaptive toggle number compensation techniques may provide significant savings in terms of power consumption of a DAC.

Adaptive toggle number compensation for reducing data dependent supply noise in digital-to-analog converters

A multi-path dual-switch DAC refers to implementing multiple paths in a switch driver and only two switches in a DAC stack of a DAC unit. In addition to multiple paths configured to improve the driving ability of the input signals, the switch driver of a multi-path dual-switch DAC unit includes two or more logic gates configured to act as multiplexers combining some of the output signals from different paths. The use of such logic gates enables using only two switches in the DAC stack unit to receive the data. Furthermore, optionally, additional logic gates may be used to combine some other output signals from different paths to generate dummy signals, thus providing internal dummy logic. The multi-path dual-switch DACs described herein may advantageously use half-clock rate and reduce or eliminate supply modulation issues, while also reducing power consumption and improving linearity compared to traditional DAC architectures.

Multi-path dual-switch digital-to-analog converter

An analog-to-digital converter (ADC) is a device that can include a reference shuffler and a loop filter. An ADC can achieve better performance with incremental adjustment of a pointer of the reference shuffler, changing coefficients of the loop filter, and storing calibration codes of the ADC in a non-volatile memory. By incrementally adjusting a pointer of the reference shuffler, a calibration can be performed more efficiently than with a random adjustment of the pointer. By temporarily changing the loop filter coefficients, a greater amount of activity can be introduced into the loop filter. This activity can allow the calibration to proceed more efficiently. By storing the calibration codes in a non-volatile memory, a search space for calibration codes can be reduced. Thus, a calibration can occur more quickly, and the calibration itself can be improved.

Flash analog-to-digital converter calibration

Analog-to-digital converters (ADCs) can be used inside ADC architectures, such as delta-sigma ADCs. The error in such internal ADCs can degrade performance. To calibrate the errors in an internal ADC, comparator offsets of the internal ADC can be estimated by computing a mean of each comparator of the internal ADC. Relative differences in the computed means serves as estimates for comparator offsets. If signal paths in the internal ADC are shuffled, the estimation of comparator offsets can be performed in the background without interrupting normal operation. Shuffling of signal paths may introduce systematic measurement errors, which can be measured and reversed to improve the estimation of comparator offsets.

Background flash offset calibration in continuous-time delta-sigma ADCS

For continuous-time multi-stage noise shaping analog to digital converters (CT MASH ADCs), quantization noise cancellation often requires estimation of transfer functions, e.g., a noise transfer function of the front end modulator. To estimate the noise transfer function, a dither signal can be injected in the front end modulator. However, it is not trivial how the dither signal can be injected, since the dither signal can potentially leak to the back end modulator and cause overall noise degradation. To address some of these issues, the dither signal is injected post the flash analog to digital converter (ADC) of the front end modulator. Furthermore, dummy comparator structures can be used to synchronize the dither with the quantization noise of the targeted flash ADC.

Dither injection for continuous-time MASH ADCS

For continuous-time multi-stage noise shaping analog-to-digital converters (CT MASH ADCs), quantization noise cancellation often requires accurate estimation of transfer functions, e.g., a noise transfer function of the front end modulator and a signal transfer function of the back end modulator. To provide quantization noise cancellation, digital quantization noise cancellation filters adaptively tracks transfer function variations due to integrator gain errors, flash-to-DAC timing errors, as well as the inter-stage gain and timing errors. Tracking the transfer functions is performed through the direct cross-correlation between the injected maximum length linear feedback shift registers (LFSR) sequence and modulator outputs and then corrects these non-ideal effects by accurately modeling the transfer functions with programmable finite impulse response (PFIR) filters.

Adaptive digital quantization noise cancellation filters for mash ADCs

For analog-to-digital converters (ADCs) which utilize a feedback digital-to-analog converter (DAC) for conversion, the final analog output can be affected or distorted by errors of the feedback DAC. A digital measurement technique can be implemented to determine timing mismatch error for the feedback DAC in a continuous-time delta-sigma modulator (CTDSM) or in a continuous-time pipeline modulator. The methodology utilizes cross-correlation of each DAC unit elements (UEs) output to the entire modulator output to measure its timing mismatch error respectively. Specifically, the timing mismatch error is estimated using a ratio based on a peak value and a value for the next tap in the cross-correlation function. The obtained errors can be stored in a look-up table and fully corrected in digital domain or analog domain.

Digital measurement of DAC timing mismatch error

For analog-to-digital converters (ADCs) which utilize a feedback digital-to-analog converter (DAC) for conversion, the final analog output can be affected or distorted by errors of the feedback DAC. A digital measurement technique can be implemented to determine switching mismatch error for the feedback DAC in a continuous-time delta-sigma modulator (CTDSM) or in a continuous-time pipeline modulator. The methodology forces each DAC unit elements (UEs) to switch a certain amount times and then use the modulator itself to measure the errors caused by those switching activities respectively. The obtained errors can be stored in a look-up table and fully corrected in digital domain or analog domain.

Digital measurement of DAC switching mismatch error

The present disclosure describes a mechanism to digitally correct for the static mismatch of the digital-to-analog converter (DAC) in at least the first-stage of a multi-stage noise shaping (MASH) analog-to-digital converter (ADC). The correction is applicable to continuous-time implementations, and is especially attractive for high-speed applications.

Cancellation of feedback digital-to-analog converter errors in multi-stage delta-sigma analog-to-digital converters

Digital-to-analog converters (DACs) are used widely in electronics. The DACs are usually not ideal and typically exhibits errors, e.g., static mismatch errors. This disclosure describes a digital calibration technique for DAC static mismatch in continuous-time delta-sigma modulators (CTDSMs). The methodology utilizes the DAC unit elements (UEs) themselves to measure each other's mismatch. There are no extra circuitries except for the logic design inside DAC drivers or comparators. The methodology is an attractive calibration technique for high performance CTDSMs, especially for high speed system in multi-gigahertz range with low over-sampling rate (OSR).

Growing community of inventors

Beijing, China

Jialin Zhao