Inside the MEMS Microphone: From Sound Wave to Digital Signal

1.Introduction
2.How It Works
3.Hardware Structure
4.Bestar's Acoustic Engineering Capabilities
5.Conclusion

Introduction
For decades, the electret condenser microphone (ECM) was the default replicator for the subject of consumer audio. ECMs are low cost, simple and functional. But they have a fundamental limitation, they're built like analog components by hand, and as such they're inconsistent from unit to unit, sensitive to heat, and hard to shrink into a cell phone.
MEMS microphones changed once and for all. MEMS(Micro Electro Mechanical Systems) are devices that feature mechanical devices in addition to electronic circuits on the same silicon chip, created by the same semiconductor processes that are used in making processors and memory. Applied to acoustic sensing this means that a microphone can be made which acts less like a traditional electromechanical and more like a precision semiconductor device.
The practical benefits are enormous. First, mems microphones​ have consistency, MEMS microphones making the process uses silicon wafers through photolithography, so among millions of units of resonant microphone, the sensitivity and the frequency response are strictly controlled. Second, thermal stability, MEMS microphones scale the reflow soldering temperatures of SMT assembly lines, a temperature at which conventional ECMs are destroyed. Third, MEMS packages regularly measure 2.5 x 1.8mm and smaller which make the ultra-thin smartphones TWS earbuds, smart vehicles and IoT devices that are synonymous with modern consumer electronics possible.
These properties have made MEMS microphones the standard for any application in which audio quality, manufacturing reliability or device miniaturization is a priority.

How It Works: Turning the Sound into Electrical Signal
A MEMS microphone works by the principle of variation of capacitance. To understand this mechanism one needs only the simplest of physics.
A capacitor is an electrical store which keeps electrical charge between two conductive plates separated by a gap. The capacitance(the amount of charge packed into it)is inversely proportional to the distance between the plates. When that distance is changed, the capacitance is changed. When a change in capacitance takes place in a charged system, there is also a change in voltage in that system. That change in voltage represents the electrical signal.
In the case of MEMS microphone, the two "plates" are the diaphragm and the backplate. The diaphragm is a thin and flexible silicon membrane, located on the back of the backplate is a rigid, perforated electrode a few microns behind it. Sound waves(a pressure wave in air)push against the diaphragm and cause it to flex. That flexing changes the gap between the diaphragm and backplate, and this changes the capacitance, and it's that which works out a voltage signal that corresponds to the sound pressure being transmitted there.
The signal generated is extremely small, in the order of microvolts. It cannot pass over a long distance without being amplified and conditioned. This is the working of the ASIC.
The ASIC (Application Specific Integrated Circuit) is the second silicon chip that is found within every MEMS microphone package. It performs three functions. First, it is what supplies a stable bias voltage to the capacitive element (called a charge pump, an internal circuit that generates a DC level of polarization voltage, in order to achieve a constant electric field across the capacitor). Second, it is a means of impedance conversion, which changes the high impedance output of the capacitive element into the signal chain driving impedance of the low impedance signal. Third, it amplifies and in digital versions, takes it and puts it in the format of a standard signal.

Hardware Structure: The Micromechanics of Semiconductor Grade
The MEMS Chip (Sensing Chip)
The piston is the element of movement. It is usually a circular or rectangular membrane made of silicon, a few micrometers thick, which is anchored at the edges and left free to flex from the middle. It has a stiffness and mass, which determine the sensitivity and frequency response characteristics of the microphone. Thinner, larger diaphragms have greater sensitivity but have less rigidity.
The rear surface on the backplate is the stationary electrode. It is pierced with an array of acoustic holes, small enough to ensure structural rigidity but big enough to ensure airflow, through which no viscous resistance prevents the diaphragm from moving. The gap between the diaphragm and backplate is normally 1-4 micrometer. Maintaining this dimension throughout the production process is one of the issues of manufacturing MEMS acoustic devices.
The Signal Processing Chip (ASIC Chip)
The ASIC performs impedance transformation, pre amplification & analog to digital conversion. In case of analog output devices, it provides a voltage signal, either single or differential with fixed gain. In digital output devices, it contains the S-D modulator which changes the analog signal to PDM (Pulse Density Modulation) or I²S bitstream.
Packaging & Acoustics Cavity
The two chips(MEMS and ASIC) are mounted inside a surface mount package, usually a metal-lid LCC housing, or plastic LGA type housing. The acoustic port is either on the bottom of the package (bottom port) or the top port.
Bottom-port microphones align the acoustic opening with a hole in the PCB below and draw sound from below the PCB. Top-port microphones are opening to the component side and receive their sound from above. The choice is dependent on the geometry of the enclosure used, the acoustic sealing requirement and the direction of the target sound source.
The ratio of front cavity volume to back cavity volume (the spaces on either side of the diaphragm) has a direct effect on sensitivity and low frequency response. And a larger front cavity will generally improve low frequency extension.

Bestar's Acoustic Engineering Capabilities
Bestar has developed its MEMS microphone portfolio through sustained investment in acoustic research and semiconductor-level process control. Beyond supplying components, Bestar can also provide solutions.

Conclusion
A MEMS microphone is a very remarkable device. This transforms variations in air pressure (sound) in to electrical information with extraordinary fidelity back to the computer, in a package smaller than a grain of rice.
This is possible because acoustics physics and semiconductor manufactory have converged. The variable capacitor principle is known more than a hundred years. What changed is being able to do that capacitor, including diaphragm, backplate, gap with a cost that does not include a level of micro-current precision.
Looking further ahead, it is the microphones that use microwatts of power, are constantly on and wake up a device only when specific key word is received. on-AI voice processing microservices looks exciting. This requires even more advanced technologies in low-power design of ASICs and in sensor fusion. MEMS microphones are well positioned for this transition, their efficiency and combining density makes them the natural basis of the next functioning of voice first gadgets.
Whether you are ordering an off-the-shelf component or would like to receive advice on acoustic system setup of any kind, please contact Bestar, Bestar is here to assist with your project from initial specification, through to production readiness.