Micro Electromechanical System (MEMS) microphones are capacitive sensors used for audio acquisition. The possibility of integration of the MEMS transducer with the CMOS read-out interface circuit in the same package, together with the small power consumption make MEMS microphones ideal for use in portable consumer electronics such as smartphones, laptops, tablets, headphones and smartwatches. Advanced features of these devices required by the market, such as noise cancellation and voice recognition, place increasing demands on performance of the microphones. This PhD thesis presents techniques to improve the Signal-to-Noise Ratio (SNR) of the interface circuit for the MEMS transducer. An electrical model of the MEMS transducer is developed in the first part of the thesis. This model includes nonlinear effects of the MEMS transducer, such as distortion. An overview of the interface circuits for MEMS capacitive transducers is reported in the second part. Direct Current (DC) biased interfaces, namely source followers, super source followers, common source amplifiers with capacitive feedback, common source amplifiers with microphone in the feedback, differential amplifiers, differential difference amplifiers and current feedback amplifiers together with voltage control Alternating Current (AC) basing and current control AC biasing are compared in terms of SNR, Acoustic Overload Point (AOP) and current consumption. Flicker noise is identified as the limiting factor for the SNR and the source follower circuit is chosen as the best candidate for improvement. In the final part of the thesis, two techniques to increase the SNR are developed: bulk biasing and dynamic capacitive attenuation. Circuits exploiting these techniques are introduced and two test-chips are fabricated. Each of the techniques improves the SNR by several , while AOP is maintained and power consumption only slightly increased.

Micro Electromechanical System (MEMS) microphones are capacitive sensors used for audio acquisition. The possibility of integration of the MEMS transducer with the CMOS read-out interface circuit in the same package, together with the small power consumption make MEMS microphones ideal for use in portable consumer electronics such as smartphones, laptops, tablets, headphones and smartwatches. Advanced features of these devices required by the market, such as noise cancellation and voice recognition, place increasing demands on performance of the microphones. This PhD thesis presents techniques to improve the Signal-to-Noise Ratio (SNR) of the interface circuit for the MEMS transducer. An electrical model of the MEMS transducer is developed in the first part of the thesis. This model includes nonlinear effects of the MEMS transducer, such as distortion. An overview of the interface circuits for MEMS capacitive transducers is reported in the second part. Direct Current (DC) biased interfaces, namely source followers, super source followers, common source amplifiers with capacitive feedback, common source amplifiers with microphone in the feedback, differential amplifiers, differential difference amplifiers and current feedback amplifiers together with voltage control Alternating Current (AC) basing and current control AC biasing are compared in terms of SNR, Acoustic Overload Point (AOP) and current consumption. Flicker noise is identified as the limiting factor for the SNR and the source follower circuit is chosen as the best candidate for improvement. In the final part of the thesis, two techniques to increase the SNR are developed: bulk biasing and dynamic capacitive attenuation. Circuits exploiting these techniques are introduced and two test-chips are fabricated. Each of the techniques improves the SNR by several , while AOP is maintained and power consumption only slightly increased.

ASIC Design Techniques and Signal Processing Supporting Power and Area Efficient High Dynamic Range Miniature Capacitive Microphones

GASPAR, SAMUEL
2025

Abstract

Micro Electromechanical System (MEMS) microphones are capacitive sensors used for audio acquisition. The possibility of integration of the MEMS transducer with the CMOS read-out interface circuit in the same package, together with the small power consumption make MEMS microphones ideal for use in portable consumer electronics such as smartphones, laptops, tablets, headphones and smartwatches. Advanced features of these devices required by the market, such as noise cancellation and voice recognition, place increasing demands on performance of the microphones. This PhD thesis presents techniques to improve the Signal-to-Noise Ratio (SNR) of the interface circuit for the MEMS transducer. An electrical model of the MEMS transducer is developed in the first part of the thesis. This model includes nonlinear effects of the MEMS transducer, such as distortion. An overview of the interface circuits for MEMS capacitive transducers is reported in the second part. Direct Current (DC) biased interfaces, namely source followers, super source followers, common source amplifiers with capacitive feedback, common source amplifiers with microphone in the feedback, differential amplifiers, differential difference amplifiers and current feedback amplifiers together with voltage control Alternating Current (AC) basing and current control AC biasing are compared in terms of SNR, Acoustic Overload Point (AOP) and current consumption. Flicker noise is identified as the limiting factor for the SNR and the source follower circuit is chosen as the best candidate for improvement. In the final part of the thesis, two techniques to increase the SNR are developed: bulk biasing and dynamic capacitive attenuation. Circuits exploiting these techniques are introduced and two test-chips are fabricated. Each of the techniques improves the SNR by several , while AOP is maintained and power consumption only slightly increased.
21-gen-2025
Inglese
Micro Electromechanical System (MEMS) microphones are capacitive sensors used for audio acquisition. The possibility of integration of the MEMS transducer with the CMOS read-out interface circuit in the same package, together with the small power consumption make MEMS microphones ideal for use in portable consumer electronics such as smartphones, laptops, tablets, headphones and smartwatches. Advanced features of these devices required by the market, such as noise cancellation and voice recognition, place increasing demands on performance of the microphones. This PhD thesis presents techniques to improve the Signal-to-Noise Ratio (SNR) of the interface circuit for the MEMS transducer. An electrical model of the MEMS transducer is developed in the first part of the thesis. This model includes nonlinear effects of the MEMS transducer, such as distortion. An overview of the interface circuits for MEMS capacitive transducers is reported in the second part. Direct Current (DC) biased interfaces, namely source followers, super source followers, common source amplifiers with capacitive feedback, common source amplifiers with microphone in the feedback, differential amplifiers, differential difference amplifiers and current feedback amplifiers together with voltage control Alternating Current (AC) basing and current control AC biasing are compared in terms of SNR, Acoustic Overload Point (AOP) and current consumption. Flicker noise is identified as the limiting factor for the SNR and the source follower circuit is chosen as the best candidate for improvement. In the final part of the thesis, two techniques to increase the SNR are developed: bulk biasing and dynamic capacitive attenuation. Circuits exploiting these techniques are introduced and two test-chips are fabricated. Each of the techniques improves the SNR by several , while AOP is maintained and power consumption only slightly increased.
MALCOVATI, PIERO
Università degli studi di Pavia
File in questo prodotto:
File Dimensione Formato  
Thesis_final_SG.pdf

embargo fino al 02/08/2026

Dimensione 5.72 MB
Formato Adobe PDF
5.72 MB Adobe PDF

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14242/188348
Il codice NBN di questa tesi è URN:NBN:IT:UNIPV-188348