Fully integrated triple-mode sigma-delta modulator for speech codec

Zou, Lei; Blasi, Marco; Rocca, Gino; Grassi, M.; Malcovati, P.; Basçhirotto, A.

doi:10.1109/norchip.2016.7792889

Cited by 3 publications

(1 citation statement)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is mainly used to vary the features of the sensor [27], [30], [75]. Sensors can also be equipped with artificial intelligence capabilities to be able to work in different situations, requiring reconfigurable ADCs [76]. Digital microphones, which include two different Σ∆Ms, are a particular point of interest.…”

Section: Introductionmentioning

confidence: 99%

Idle Tones Reduction in Digital Single-Bit ΣΔ Modulators

Vera

Patón

Straeussnigg

2019

2019 IEEE International Symposium on Circuits and Systems (ISCAS)

View full text Add to dashboard Cite

Analog-to-digital converters based on Discrete-Time Sigma-Delta modulators have enjoyed great popularity in recent decades, mainly due to their high performance in the bandwidth required by sensors. This type of converter has the highest reported figure-ofmerit values for audio sensors, and they have a high tolerance to process, supply voltage, and temperature variations. They have reduced sensitivity to clock jitter and to errors of the digital-to-analog converter. For these reasons, they are widely used in applications requiring high robustness. These types of converters are present in a large number of commercial products and have been extensively studied from an academic point of view due to their economic potential. This thesis proposes the development of Sigma-Delta modulators with architectures optimized for digital MEMS microphones, focusing on both the analog-to-digital converter and the output data interface.Data acquisition circuits for digital microphones have two distinct blocks: the analogto-digital converter and the standardized output interface, which generates a binary encoded signal to be sent to an audio codec. Sigma-Delta modulators are used in both blocks but they require different optimizations. The analog-to-digital converter generates a multi-bit signal while the output interface generates a single-bit signal.The first block of the sensor is associated with the analog-to-digital conversion, which translates the variation of the analog input signal into a digital multi-bit signal. The input signal to the analog-to-digital converter is generated by a preamplifier, which amplifies the signal generated by the capacitive MEMS microphone. Undesired signals can be present at the analog-to-digital converter input, such as low-frequency signals, that increase the noise power of the converter because the modulator is forced to toggle between three adjacent levels. Different error sources contribute to the noise power increase: digitalto-analog converter mismatch, loop filter linearity error, and quantization error. These error sources are analyzed, and estimation equations are proposed by means of deterministic and stochastic methods. These equations can be used to optimize the modulator parameters in order to reduce such noise power increments. Another situation discussed in this dissertation derives from the high sensitivity peaks that MEMS microphones have in the ultrasonic range. These peaks appear because the MEMS sensor behaves as a res-6.

show abstract

Section: Introductionmentioning

confidence: 99%