Micro-electro-mechanical system integration and performance, cost and reliability of the comprehensive consideration
Whether the integration of MEMS with the controversial nature of integration is to achieve high-performance, low-cost, high reliability and high quality the only way, but its large R & D and production inputs needed to support a larger volume, but also a viable process to resolve. This paper analyzes the integrated and non-integration of two kinds of perspectives and their reality based on our vision MEMS integration trends.
Integrated micro-electromechanical systems (MEMS) as a viable approach? When may require additional time and take additional risks, MEMS single-chip integration of the project should be chosen approach? Will probably be a few die under the same package is a better solution, but should be used what standards? This question is not easy to answer, because there are some uncertain factors, and often only in the product to the market for quite a long time before finally determining its success. MEMS integration is an early product life cycle cost of the investment return on investment and the latter part of a compromise between, but often by the impact of technology and application requirements. In most cases, if the sensors do not have to because of specific applications must be separated with the electronic circuit, the single-chip integration is the best solution, but in the chemical sensors and high temperature harsh environments such as sensors and electronic circuits must be separated. Perhaps just as in the past 25 years, the development of semiconductor micro-electronics technology, as, when the tools, technology and production experience at a lower risk to circumvent these uncertainties, the integration will become a necessity.
Like most engineering solutions, the success of MEMS products to resolve design and development, and trade-offs and reduce the cost of production-related issues. And extend the learning cycle, complex equipment and technology related to time risk, making the field of integration in micro-machining is not common. But it is interesting is that in the microelectronics industry appears to always have such a belief, that as long as efforts continued to be able to get a breakthrough.
To bring the advantages of an integrated
When the IC chip has only a few transistors or at an early stage CMOS?? S technology, semiconductor technology has a number of uncertain variables, these variables are more like a random number instead of the normal function. Early tools used to define the planar device is only a few simple process maps and models, the device characteristic parameters are usually mathematical approximations and slide rules (slide-rule) to establish. At that time, people are based solely on yield and defect density, as evidenced by 100 million transistor IC can not be achieved. Looking back on our use of 100 million transistors Itanium processor chip, can be found lead us forward along the Moore's Law, and has 200 billion U.S. dollars to form an industry, advances in technology to reduce risk of failure in understanding the design and manufacture of tools on the progress Results.
IC single-chip integration achieved by the best engineering solution seems to have resolved most of the controversy over the past few years, high performance, low cost, reliability and quality level addition to an integrated approach to things can not be achieved, integration has become the current IC manufacturers all the key business differentiation. In fact as long as we observe can be found that even with the use of bipolar integrated CMOS process and mixed-signal CMOS process devices in real life has been a lot of applications.
Why should we integrate MEMS?
The problem with a similar problem on the IC integration. The primary factor to consider is whether there is a large enough volume to amortize the integration brought about by the development and processing costs, in addition to possible technological problems must be addressed. The most crucial argument is there any other ways to achieve the required volume of low-cost goal. When the physical size limitations can not be used due to other possible methods, often make use of the integrated final decision. Large number of products quality and reliability levels that may affect this trade-off decisions, while ADI's from the launch of more than 150 million pieces on the single-chip MEMS devices derived from empirical evidence, single-chip integrated MEMS are usually able to achieve less than 1ppm quality levels, and 10 billion hours of mean time between failures (MTBF).
Solution from a technical point of view, all the components are placed on a chip has some obvious advantages: high noise conditions, the small-signal can be subjected to minimal stress, electromagnetic interference, parasitic capacitance and leakage current, etc. "unknown variable" impact. Use of common IC design techniques (for example, lead in technology, cross-space (cross-quading)) and the switched-capacitor charge management can eliminate the temperature and other unknown factors from the impact. As the integrated circuit will have low-level signals together, while the thermal environment factors remain unchanged, and thus can get a lower thermal hysteresis, and better connected features. The need to address the millions of display unit and the video-frequency excitation applications, the size limitations and interconnect density, demonstrated using an integrated MEMS device is correct. When the hand-held devices or medical products require its thickness or width shall not exceed the limits of the system package, the same proof of an integrated MEMS device is reasonable.
Address these problems another way is to generate a larger signal, but usually they require a larger silicon area, higher power and lower impedance, and a multi-chip solutions in the form to achieve. From the real environment to obtain sufficient information, it is a set of sensors to increase the MEMS components, and in the remote location of the signal to compensate. As a result, interconnect cost, size and complexity of non-linear (exponential) growth, resulting in high costs and difficult components of the packaging and testing. ADI's single-chip gyroscope (Gyro, see Figure 1), experience has shown that, compared with the off-chip processing, on-chip integration brings the advantage of handling the signal amplitude of at least two orders of magnitude less signal. The silicon sensors in general need to increase the size of 10-1000 times that can only produce the same signal to noise ratio, which will increase stress management (stressmanagement) and long-term stability of the major design challenges.
Integration is not always the best choice
Integration of MEMS devices is not certain the best solution. Integrated MEMS products may require longer than non-integrated design development time, especially when related to the development process at the same time when you need to. MEMS Design and Technology because of the close link between the completion of the building several different designs and fully understand its production yield ago, it is difficult to determine a feasible process. As the system-level development cycle problem, the customer may be due to appraisal cycle or system-level quality requirement was declared "not to make any change."
As in recent years, there is no large volume, so MEMS integration may be bad economics. Annually from 100,000 to save one U.S. dollars on the components can not be offset by spending 10 U.S. dollars on each component of the development costs.
From a technical and economic point of view, non-integrated solution may be the best product design, especially when the non-integrated options would be enough time performance. In addition, the larger wafers is sometimes necessary to meet the external environmental conditions. Embedded strain gauge (embeddedstraingauge) or a capacitive diaphragm (capacitivediaphragm) configured to create a simple mechanical process can be compared with IC plus the cost of lower MEMS process. A combination of physical adjustments and different combinations of materials and processes Radio Frequency modules and optical components are suitable for certain non-integrated MEMS devices. These non-integrated version is often due to its higher performance or enhance the special characteristics of the system have higher profit margins and higher price.
MEMS technology is not static, take a look at the current widespread use of three U.S. dollars accelerometers or is likely to form a new display era megapixels MEMSDLP chips, can be found integrated MEMS has an important economic advantage.
Connection is to promote communication and increase the complexity of MEMS devices and functions is a strong force. The next generation of automotive safety systems sensors are being developed using the two-wire interface, which a single line in the same data and power transmission create a serial configuration, with e-diagnostics and calibration functions (the implementation of all standard formats) of the total two-way. Automotive Safety chip market more than 100 million annually, sensor interface, a new era is approaching. If the combined ZigbeeAlliance the company's wireless network, using standard protocols to communicate the potential of MEMS devices is to promote greater market volume and higher integration.
Built on the deep reactive ion etching (DRIE) process on the process option to make it possible to complete in the main standard IC process was to create MEMS structures. When combined have a buried oxide layer (buriedoxide) of the SOI wafer, the integration of MEMS will be the mainstream of IC foundry process. Buried oxide layer would be likely to die on the MEMS device from the sub-micron CMOS process surrounded released, as shown in Figure 2.
MEMS devices and integrated circuits have the same trends, because they have the same processing tools and processes. When the need to increase the functionality and significantly lower prices, while not sacrificing quality and reliability when the integration is the only option.
Radio Frequency Articles
- The use of semiconductor technology of information between doctors and patients management
- Micro-electro-mechanical system integration and performance, cost and reliability of the comprehensive consideration
- ZigBee positioning solution
- Advanced RFID reader applications processor requirements
- Advanced RFID reader applications processor requirements
- Wireless sensor networks based on the prison security system design
- Wireless IPPBX System Design and Implementation of
- Based on the MCF5213 and Zigbee technology, wireless intercom systems
- NI PXI-based radio receiver technology, technical brief
- WLAN Positioning System Analysis and Design
- Wireless SoC for signal integrity analysis
- 4G air interface generic OFDMA and MIMO technology to explore
- ZigBee technology to increase the use of piggyback method of network performance
- Lighting control for the communication protocol
- RFID system design configuration management software
- Beamforming based on the electronic tracking technology research
- WiMAX technology based on the 5.8G RF Wireless Network System Design
- Multi-function monitor wireless communication systems
- RF Crystal Oscillator System Design and Analysis
- Electrostatic control of the RF tag solutions
Can't Find What You're Looking For?
Rating: Not yet rated