Design and implementation of the hottest CMOS capa

  • Detail

Design and implementation of CMOS capacitive micro microphone

with the rise of intelligence, people pay more and more attention to the demand for sound quality and lightness. In recent years, the widely used noise suppression and echo cancellation technology is to improve the sound quality. Compared with the traditional electret microphone (ECM), the capacitive MEMS microphone is made of silicon semiconductor material, which facilitates the integration of analog amplification circuit and ADC (∑)- Δ ADC) circuit, realizing analog or digital MEMS microphone components, and manufacturing miniaturized components, which are very suitable for light, thin and short portable devices. This paper will introduce the design and manufacture of CMOS MEMS microphone, and compare the difference between pure MEMS and CMOS process micro induction microphone

principle of capacitive microphone

mems microphone is a kind of miniature sensor. The principle is to use the pressure gradient generated by the sound change to make the acoustic diaphragm of the capacitive microphone deform due to the sound pressure interference, and then change the capacitance between the acoustic diaphragm and the silicon back plate. The change of the capacitance value is transformed from the capacitance voltage conversion circuit into the output change of the voltage value that private enterprises have focused on the research and development of aluminum alloy products for 30 years. Then, the MEMS sensor is generated by the amplification circuit to obtain the voltage amplification output, so as to convert the sound pressure signal into a voltage signal. Here, a high impedance resistor must be used to provide a bias voltage Vpp for the MEMS sensor, so as to generate a fixed charge on the MEMS sensor. The final output voltage will be related to the deformation of the VPP and the diaphragm Δ D is proportional. The deformation of the diaphragm is related to its rigidity. The lower the rigidity, the greater the deformation; On the other hand, the output voltage is inversely proportional to D (air gap), so the lower the air gap, the better the output voltage and sensitivity. However, this will be limited by the pull in voltage of MEMS sensor, that is, the maximum limit value of electrostatic field of MEMS sensor

cmos MEMS microphone circuit design

in the design of CMOS micro microphone, the circuit is a very important link, which will affect the operation, sensing and system sensitivity of the micro microphone. The induced charge of the electret condenser microphone is generated by the electret charge provided by the electret material itself, while the condensing condenser microphone provides a fixed charge source by extracting a bias voltage from the CMOS operating voltage and supplying it to the acoustic diaphragm of the microphone through a high impedance resistor. At this time, if the acoustic diaphragm is driven by sound pressure to produce displacement change, the voltage of the electrode plate (sensing end) will change. Finally, the circuit design of the analog microphone can be realized by amplifying the signal through the circuit amplifier; If you add a sigma- Δ ADC analog-to-digital conversion circuit can complete the circuit design of digital microphone (generally, the output signal of digital microphone is 1-bit PDM output)

cmos MEMS microphone process classification

from the perspective of MEMS microphone manufacturing, MEMS components of integrated CMOS circuits can be divided into three types at the current technical level. Pre CMOS MEMS process: first fabricate MEMS structure, and then fabricate CMOS components; Intra CMOS MEMS process: mixed manufacturing of CMOS and MEMS component processes; Post CMOS MEMS process: first realize CMOS components, and then manufacture MEMS structures. Generally speaking, the first two methods cannot be carried out in traditional wafer factories, while post CMOS MEMS can be produced in semiconductor wafer foundries

the following figure briefly describes the manufacturing method of post CMOS MEMS. In the post CMOS MEMS process, special attention should be paid to that additional heat treatment or high-temperature process should not affect the physical properties of CMOS components and the stress state of MEMS, so as not to affect the initial stress of the diaphragm. Xinchuang Technology Co., Ltd. has overcome many technical problems and completely adopted standard CMOS technology to manufacture circuit components and MEMS microphone structures at the same time. After the CMOS part is completed, grind the back of the chip to an appropriate thickness to meet the packaging requirements. Finally, hydrofluoric acid solution (HF) was used to remove sacrificial oxides to release the suspended structure. In addition, the etching method that can completely remove the sacrificial materials without damaging the microphone diaphragm shall be considered in the design, and the adhesion between the microphone diaphragm and the back electrode plate shall be avoided

stickiness: since the distance between the microphone diaphragm and the back electrode plate is only a few microns, under this size, when the surface tension, van der Waals force, electrostatic force, ionic bond and other forces are greater than the restoring force of the microphone diaphragm, the microphone diaphragm will be permanently deformed and attached to the back electrode plate, so vibration cannot be generated. Generally, the main causes of adhesion of MEMS suspension structure can be divided into two types: the first type occurs after the microphone diaphragm is released, and the microphone diaphragm is affected by surface tension, so it is pulled close to the back electrode plate. If the surface forces such as van der Waals force or hydrogen bond force are greater than the restoring force of the microphone diaphragm at this time, the structure will produce adhesion and cannot be restored; The second type is that when the suspended structure is impacted by external force or attracted by electrostatic force in use and falls into the area where the surface force is greater than the restoring force, the adhesion phenomenon will also occur. Therefore, in the structural design, the structural deformation of microphone diaphragm after release must be specially considered, and the important structural parts must be strengthened, and the special design should be used to reduce the occurrence of adhesion

differences between pure MEMS and CMOS processes

MEMS Micro microphones developed by most enterprises are mainly divided into two forms: the first is to avoid oil overflow, use professional MEMS OEM to manufacture MEMS IC, and add an ASIC amplifier to package MEMS IC and ASIC IC into MEMS microphone chip by SIP packaging. This part must protect the diaphragm from damage during IC packaging, and its packaging cost is relatively high; The other is to use the CMOS wafer factory to manufacture the ASIC part, and then use the process of more than 410 years to form the structural part of MEMS. It seems that the MEMS process technology can not be completed in the standard CMOS wafer factory at present. This is mainly because the diaphragm needs to deposit polymer materials, which have not been used in the current standard semiconductor IC process. In addition, after the CMOS process is completed, the diaphragm should be etched on the front of the chip and the cavity and acoustic hole should be etched on the back of the chip. This step is completed by carrier wafers, which have not yet been created in standard CMOS foundries

at present, the biggest issue is how to break through the packaging technology of these two forms of MEMS microphones. Its patents are controlled by micro microphone enterprises in the United States. Therefore, MEMS microphone market share is mainly distributed in the hands of a few enterprises

Xinchuang technology adopts the method of forming cavity and acoustic hole from the back of the chip as the release of MEMS structure after the completion of CMOS process. This part can be completed in the existing CMOS wafer factory without using special machines and materials, thus reducing the development cost. In addition, the products of Xinchuang technology can directly use wafer level packaging technology to integrate CMOS circuits and micro microphones on the same chip, which can also avoid damaging the diaphragm during the packaging process

mems microphones have now replaced ECM microphones which are extremely sensitive to experimental risks and are widely used in (especially intelligent). The main reason is that MEMS microphones have the advantages of good weather resistance, small size and easy digitization. MEMS microphone is made of semiconductor material, which has stable characteristics and will not be affected by ambient temperature and humidity, so it can maintain stable sound quality. When the electronic products are assembled in the tin furnace, the temperature is as high as 260 ℃, which often destroys the diaphragm of the ECM microphone and must be reworked, which will increase the additional cost. The MEMS microphone will not affect the material because of the high temperature of the tin furnace, which is suitable for the automatic assembly of SMT. After the microphone signal is digitized, it can be subjected to signal processing such as de-noising, sound bunching and echo cancellation, so as to provide excellent call quality. At present, digital technology has been adopted in many intelligent products, and there are signs of accelerated adoption in functions. In addition, notebook computers are also the mainstream using MEMS microphones, and set-top box manufacturers are also actively trying to apply MEMS microphones to the development of voice controlled set-top boxes

Copyright © 2011 JIN SHI