Micro-Electro-Mechanical Systems (MEMS) have become a central technology platform for miniaturised devices in sensing, timing, and actuation, owing to the possibility of integrating mechanical structures, electrical interfaces and signal-conditioning electronics within compact, batch-fabricated systems. The extension of this paradigm to loudspeakers is a natural objective for earphones, hearing aids and other space-constrained audio applications, where MEMS devices promise compact footprint, reduced assembly complexity and compatibility with standard semiconductor fabrication and packaging technologies compared to conventional voice-coil microspeakers. The limited diaphragm area and small achievable mechanical displacement, however, make low-frequency sound generation particularly challenging in free-field conditions. Therefore, most of the devices reported in the literature target in-ear configurations, where the small enclosed volume of the ear canal allows achieving higher sound pressure levels with moderate diaphragm motion. Among the various actuation principles proposed at the microscale, piezoelectric transduction has emerged as the most mature: state-of-the-art piezoelectric MEMS speakers now match the acoustic performance of commercial in-ear products, although their adoption in mass-market devices still requires further effort to sustain such performance while ensuring full compatibility with established CMOS fabrication flows and reliable operation under arbitrary drive conditions. The present work addresses these challenges through the design, modeling and experimental characterization of MEMS loudspeakers based on novel actuation schemes. High-performance piezoelectric devices, designed within the research group prior to this work, are first modelled and experimentally characterised to establish a quantitative benchmark against which the subsequent contributions of the thesis are evaluated. The capabilities of electromagnetic and electrostatic actuation are then assessed through hypothetical designs analysed by finite-element modelling, whose results consolidate the piezoelectric route as the most promising path towards in-ear applications. On this basis, a new piezoelectric actuation scheme is introduced, in which two independent piezoelectric patches deposited on the same diaphragm are driven in anti-phase to add their respective displacement contributions. The proposed device is modelled, fabricated on a polysilicon-based wafer platform with a PZT piezoelectric layer, and experimentally tested: its prototype delivers a Sound Pressure Level (SPL) of 109.5 dB SPL from 100 Hz onward under a 30 Vpp drive, a Total Harmonic Distortion (THD) below 1% at 94 dB SPL and 1 kHz, and a compact footprint of 4.5 mm × 4.5 mm. These figures qualify it as one of the highest-performing piezoelectric MEMS loudspeakers reported in the literature, while remaining fully compliant with industry-standard piezoelectric fabrication processes. The same actuation scheme is finally extended to a set of theoretical designs aimed at refining the architecture, leading to a configuration predicted to reach 113 dB SPL from 100 Hz onward while relying on a single piezoelectric patch with a capacitance of 15 nF at 0 V, and therefore on a substantially reduced power consumption. Taken together, the devices proposed in this work represent a promising step towards high-performance piezoelectric MEMS loudspeakers that meet the requirements of in-ear consumer products while guaranteeing low power consumption, manufacturability within established fabrication flows, and robust performance across the full range of operating conditions.
I sistemi micro-elettro-meccanici (MEMS) sono diventati una piattaforma tecnologica centrale per dispositivi miniaturizzati nei settori di sensing, timing e attuazione, grazie alla possibilità di integrare strutture meccaniche, interfacce elettriche ed elettronica di condizionamento del segnale in sistemi compatti e fabbricati in batch. L'estensione di questo paradigma agli altoparlanti rappresenta un obiettivo naturale per auricolari, apparecchi acustici e altre applicazioni audio con vincoli di spazio, dove i dispositivi MEMS promettono un ingombro ridotto, una minore complessità di assemblaggio e una migliore compatibilità con le tecnologie standard di fabbricazione e packaging dei semiconduttori rispetto ai tradizionali altoparlanti a bobina mobile. La ridotta area del diaframma e il limitato spostamento meccanico ottenibile rendono tuttavia la generazione di suoni a bassa frequenza particolarmente critica in condizioni di campo libero. Per questo motivo, la maggior parte dei dispositivi presentati in letteratura è concepita per configurazioni in-ear, dove il ridotto volume racchiuso dal canale uditivo consente di raggiungere livelli di pressione sonora più elevati anche con spostamenti moderati del diaframma. Tra i diversi principi di attuazione proposti alla microscala, la trasduzione piezoelettrica si è affermata come la più matura: gli altoparlanti MEMS piezoelettrici allo stato dell'arte raggiungono ormai prestazioni acustiche paragonabili a quelle dei prodotti in-ear commerciali, sebbene la loro adozione su larga scala richieda ancora uno sforzo ulteriore per mantenere tali prestazioni garantendo al contempo piena compatibilità con i flussi di fabbricazione CMOS consolidati e un funzionamento affidabile in qualunque condizione di pilotaggio. Il presente lavoro affronta queste sfide attraverso la progettazione, la modellazione e la caratterizzazione sperimentale di altoparlanti MEMS basati su nuovi schemi di attuazione. Dispositivi piezoelettrici ad alte prestazioni, progettati all'interno del gruppo di ricerca prima dell'inizio di questo lavoro, vengono dapprima modellati e caratterizzati sperimentalmente per stabilire un benchmark quantitativo rispetto al quale valutare i successivi contributi della tesi. Le potenzialità dell'attuazione elettromagnetica ed elettrostatica vengono quindi valutate attraverso design ipotetici analizzati tramite modellazione agli elementi finiti, i cui risultati confermano la via piezoelettrica come la più promettente per le applicazioni in-ear. Su queste basi viene introdotto un nuovo schema di attuazione piezoelettrica, in cui due patch piezoelettriche indipendenti depositate sullo stesso diaframma vengono pilotate in controfase in modo da sommare i rispettivi contributi di spostamento. Il dispositivo proposto viene modellato, fabbricato su una piattaforma wafer a base di polisilicio con uno strato piezoelettrico in PZT, e caratterizzato sperimentalmente: il prototipo realizzato raggiunge un livello di pressione sonora (SPL) di 109.5 dB SPL a partire da 100 Hz con un pilotaggio di 30 Vpp, una distorsione armonica totale (THD) inferiore all'1% a 94 dB SPL e 1 kHz, e un ingombro compatto di 4.5 mm × 4.5 mm. Questi risultati lo collocano tra gli altoparlanti MEMS piezoelettrici con le migliori prestazioni riportate in letteratura, pur rimanendo pienamente compatibile con i processi di fabbricazione piezoelettrica standard del settore. Lo stesso schema di attuazione viene infine esteso a una serie di design teorici volti a raffinare l'architettura, portando a una configurazione che, secondo le previsioni, raggiunge 113 dB SPL a partire da 100 Hz pur affidandosi a una singola patch piezoelettrica con una capacità di 15 nF a 0 V, e dunque a un consumo energetico sostanzialmente ridotto. Nel complesso, i dispositivi proposti in questo lavoro rappresentano un passo promettente verso altoparlanti MEMS piezoelettrici ad alte prestazioni, in grado di soddisfare i requisiti dei prodotti in-ear di consumo garantendo al contempo un basso consumo energetico, la producibilità tramite flussi di fabbricazione consolidati, e prestazioni robuste in tutte le condizioni operative.
Design, modeling, and comparative characterization of MEMS loudspeakers with novel actuation mechanisms for in-ear applications
Perli, Filippo Pietro
2025/2026
Abstract
Micro-Electro-Mechanical Systems (MEMS) have become a central technology platform for miniaturised devices in sensing, timing, and actuation, owing to the possibility of integrating mechanical structures, electrical interfaces and signal-conditioning electronics within compact, batch-fabricated systems. The extension of this paradigm to loudspeakers is a natural objective for earphones, hearing aids and other space-constrained audio applications, where MEMS devices promise compact footprint, reduced assembly complexity and compatibility with standard semiconductor fabrication and packaging technologies compared to conventional voice-coil microspeakers. The limited diaphragm area and small achievable mechanical displacement, however, make low-frequency sound generation particularly challenging in free-field conditions. Therefore, most of the devices reported in the literature target in-ear configurations, where the small enclosed volume of the ear canal allows achieving higher sound pressure levels with moderate diaphragm motion. Among the various actuation principles proposed at the microscale, piezoelectric transduction has emerged as the most mature: state-of-the-art piezoelectric MEMS speakers now match the acoustic performance of commercial in-ear products, although their adoption in mass-market devices still requires further effort to sustain such performance while ensuring full compatibility with established CMOS fabrication flows and reliable operation under arbitrary drive conditions. The present work addresses these challenges through the design, modeling and experimental characterization of MEMS loudspeakers based on novel actuation schemes. High-performance piezoelectric devices, designed within the research group prior to this work, are first modelled and experimentally characterised to establish a quantitative benchmark against which the subsequent contributions of the thesis are evaluated. The capabilities of electromagnetic and electrostatic actuation are then assessed through hypothetical designs analysed by finite-element modelling, whose results consolidate the piezoelectric route as the most promising path towards in-ear applications. On this basis, a new piezoelectric actuation scheme is introduced, in which two independent piezoelectric patches deposited on the same diaphragm are driven in anti-phase to add their respective displacement contributions. The proposed device is modelled, fabricated on a polysilicon-based wafer platform with a PZT piezoelectric layer, and experimentally tested: its prototype delivers a Sound Pressure Level (SPL) of 109.5 dB SPL from 100 Hz onward under a 30 Vpp drive, a Total Harmonic Distortion (THD) below 1% at 94 dB SPL and 1 kHz, and a compact footprint of 4.5 mm × 4.5 mm. These figures qualify it as one of the highest-performing piezoelectric MEMS loudspeakers reported in the literature, while remaining fully compliant with industry-standard piezoelectric fabrication processes. The same actuation scheme is finally extended to a set of theoretical designs aimed at refining the architecture, leading to a configuration predicted to reach 113 dB SPL from 100 Hz onward while relying on a single piezoelectric patch with a capacitance of 15 nF at 0 V, and therefore on a substantially reduced power consumption. Taken together, the devices proposed in this work represent a promising step towards high-performance piezoelectric MEMS loudspeakers that meet the requirements of in-ear consumer products while guaranteeing low power consumption, manufacturability within established fabrication flows, and robust performance across the full range of operating conditions.| File | Dimensione | Formato | |
|---|---|---|---|
|
thesis_FilippoPietroPerli.pdf
non accessibile
Descrizione: Manoscritto tesi
Dimensione
96.9 MB
Formato
Adobe PDF
|
96.9 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/262197