Alfons Dehe

Silicon microphone based on surface and bulk micromachining

In this paper a silicon microphone which can be fabricated using standard semiconductor processes is presented. The acoustic-electrical transducer is based on the capacitance change of a movable 400 nm thin poly-silicon membrane with different diameters (800-1200 µm). A source follower was integrated to transform the impedance. The complete chip is 2×2×0.5 mm3 in size. The sensitivity achieved is in the range of 0.4 to 3.2 mV Pa-1.

Analytical analysis and finite element simulation of advanced membranes for silicon microphones

In this paper, advanced membrane designs are simulated in order to improve the sensitivity of micromachined silicon condenser microphones. Analytical analyzes and finite element simulations have been carried out to derive algebraic expressions for the mechanical compliance of corrugated membranes and membranes supported at spring elements. It is shown that the compliance of both types of membranes can be modeled with the help of an enhanced theory of circular membranes. For spring membranes, a numerically derived and design dependent constant takes into account the reduced suspension. The mechanical stress in corrugated membranes is calculated using a geometrical model and is confirmed by finite element simulations. A very good agreement between theory and experimental results is demonstrated for spring membranes of different shape and for membranes with varying number of corrugations. In a silicon microphone application, a high electro-acoustical sensitivity up to 8.2 mV/Pa/V is achieved with a membrane diameter of only 1 mm.

Silicon micromachined microphone chip at Siemens

Applications ranging from hearing aids over communication to noise cancellation open up a high volume market for low‐cost, batch producible and reliable microphones. To obey these conditions, a single‐chip capacitive microphone has been developed at Siemens, utilizing a modified standard CMOS process with adjacent bulk micromachining. In a first step, the microphone is integrated with a source follower, enabling low output impedance of the signal. The technology allows for the future integration of advanced circuitry. The microphone consists of an acoustically sensitive polycrystalline silicon membrane and a highly perforated back‐plate as the counter electrode. To achieve highly sensitive devices, special emphasis was given to the stress of the polycrystalline silicon membrane, which should be slightly tensile. Another key issue during the fabrication and in operation is to prevent stiction of the sensitive membrane. Since the overall chip size is below 3‐mm side length, surface mounting in low‐cost SMD packages is possible.

Design of a poly silicon MEMS microphone for high signal-to-noise ratio

This paper reports on the state of the art silicon micromachined microphone utilizing a dual poly silicon membrane system. MEMS chips from 1.4mm down to 1.0mm side length are applied for mobile communication. Design aspects related with key performance parameters such as sensitivity, signal to noise ration and distortion are discussed. Sensitivity of - 38BV/Pa is achieved for different microphone membrane diameters. A maximum signal to noise ration of 66dB(A) for the largest system could be achieved. The perfect fit of simulation versus measurements enables deeper analysis and balancing of noise contributors. Environmental noise suppression of 5dB by acoustical high pass design is demonstrated.

Silicon Microphones with Low Stress Membranes

A new silicon condenser microphone process for low stress diaphragm is presented. The diaphragm is formed either by a polysilicon layer or by a monocrystalline silicon layer of a silicon on insulator substrate (SOI). Acoustical sensitivities up to 8.2 mV/Pa with a 1×1 mm2 diaphragm for a bias voltage of only 1 V have been achieved. The A-weighted noise voltage was found to be 7 eV in the band of 100 Hz to 10 kHz.

Improved signal-to-noise ratio of silicon microphones by a high-impedance resistor

This paper presents the improvement of the signal-to-noise ratio of a silicon microphone by utilizing a high-impedance load resistor. A model is described which considers the acoustical and electrical parts of the microphone. Based on an equivalent circuit diagram, the model is able to simulate the sensitivity and the noise thus the signal-to-noise ratio. Describing the noise spectrum differentiated by parts, the thermal noise of the load resistor proves to be a main noise source as long as the resistance is relatively low.

Enhanced design of microsystems by combining lumped and distributed system-level models

We highlight the benefits of a tailored system-level modeling approach for the enhanced design of MEMS devices and systems demonstrated for an industrial capacitive silicon microphone. The performance of such microphones is determined by distributed effects like viscous damping and inhomogeneous capacitance variation across the membrane as well as by system-level phenomena like package-induced acoustical effects and the impact of the electronic circuitry for biasing and read-out. In order to meet these antipodal targets, a so-called mixed-level simulation approach is applied, which combines distributed and lumped element models on system-level. This provides maximum insight into the device and system operation while keeping the computational expense low. The presented model enables the investigation of the most relevant figures of merit such as the frequency response, the total harmonic distortion, and the signal-to-noise ratio as well as the analysis of their respective dependence on design parameters. The accuracy of the derived model and its potential for predictive simulation is assured by extensive calibration of each submodel and demonstrated through comparison to measured data.

Performance and noise analysis of capacitive silicon microphones using tailored system-level simulation

A fully coupled fluidic-electro-mechanical system-level model has been assembled and applied to existing and novel silicon microphone designs. Distributed and non-linear effects like fluidic damping and electrostatic forces and their impact on the overall system performance have been investigated. All relevant contributions like the package-induced acoustical effects and the electronic circuitry for biasing and read-out are included as well. Employing the fluctuation-dissipation theorem to our model we are able to predict and discriminate the noise contribution of single microphone regions in order to suggest design measures for the enhancement of the total signal-to-noise ratio (SNR) of the device. Dedicated calibration and validation of the single submodels by laser-vibrometric measurements assure the accuracy and predictive power of the presented model.

Modeling Methods of MEMS Micro-Speaker with Electrostatic Working Principle

The market for mobile devices like tablets, laptops or mobile phones is increasing rapidly. Device housings get thinner and energy efficiency is more and more important. Micro-Electro-Mechanical-System (MEMS) loudspeakers, fabricated in complementary metal oxide semiconductor (CMOS) compatible technology merge energy efficient driving technology with cost economical fabrication processes. In most cases, the fabrication of such devices within the design process is a lengthy and costly task. Therefore, the need for computer modeling tools capable of precisely simulating the multi-field interactions is increasing. The accurate modeling of such MEMS devices results in a system of coupled partial differential equations (PDEs) describing the interaction between the electric, mechanical and acoustic field. For the efficient and accurate solution we apply the Finite Element (FE) method. Thereby, we fully take the nonlinear effects into account: electrostatic force, charged moving body (loaded membrane) in an electric field, geometric nonlinearities and mechanical contact during the snap-in case between loaded membrane and stator. To efficiently handle the coupling between the mechanical and acoustic fields, we apply Mortar FE techniques, which allow different grid sizes along the coupling interface. Furthermore, we present a recently developed PML (Perfectly Matched Layer) technique, which allows limiting the acoustic computational domain even in the near field without getting spurious reflections. For computations towards the acoustic far field we us a Kirchhoff Helmholtz integral (e.g, to compute the directivity pattern). We will present simulations of a MEMS speaker system based on a single sided driving mechanism as well as an outlook on MEMS speakers using double stator systems (pull-pull-system), and discuss their efficiency (SPL) and quality (THD) towards the generated acoustic sound.

Modeling distributed electrostatic effects in silicon microphones and their impact on the performance

We present a system-level model for fast and efficient investigations of distributed electrostatic effects in state-of-the-art silicon microphones. Combining lumped and distributed submodels it accounts for electrostatic forces and capacitive read-out, including non-linearities, fringing fields and parasitics. The derived model is calibrated using electrostatic finite element (FE) simulations and validated by measurements. The non-linearities caused by electrostatic effects have a decisive impact on the sensitivity of the microphone and the distortion of the transduced acoustical signal. Hence, the proposed model provides important insights into the operation of the device, which can be employed to optimize the microphone characteristics.