Volume manufacturing
Micro-Electro-Mechanical Systems (MEMS) have been with us for some time, being promised as the next big thing for many years. One of the major stumbling blocks has been manufacturing devices at volume production. Most new MEMS devices suffer from their manufacturing needs. The semiconductor industry is set up to make silicon on a CMOS platform and most are unwilling to invest in a new tool set for every new device that is designed. When a start up company comes forward and says they have the first purely CMOS based MEMS device and claim to have all the patents tied up for MEMS manufacturing on a CMOS platform then it all sounds incredibly interesting.
The people behind Akustica have a very strong pedigree. The company is a spin off of the very successful MEMS laboratory at Carnegie Mellon University (CMU) in the USA. CMOS MEMS technology was developed at CMU starting in 1994 under DARPA sponsorship. Dr. Ken Gabriel was heading the MEMS program at DARPA. At the time, CMU's main focus was on inertial and accelerometer applications. In 1997, Dr. Gabriel left DARPA to join the faculty at CMU in Pittsburgh to become Director of the CMUMEMS Laboratory where he started a research program in acoustic and audio MEMS. The programme went on to have many successes and the work of Dr. Gabriel saw him claimed as the architect of MEMS. Dr. Gabriel pioneered the MEMS field and led the development of IC-based MEMS while working at AT&T Bell Labs.
Dr. Gabriel spent five years with DARPA, where he started and managed their MEMS program as Director of the Electronics Technology Office and was responsible for roughly half of the U.S. Government's electronics technology investments including the Carnegie Mellon project. He has also been a visiting professor for the Institute of Industrial Science at the University of Tokyo, heading joint research projects with IBM, Toyota and Ricoh, as well as a visiting scientist at the Naval Research Laboratory. He has helped found two other successful technology companies and serves on advisory boards for three others.
With such a rich history it was no surprise when Dr. Gabriel teamed up with serial entrepreneur Jim Rock in 2001 and launched Akustica with the aim of developing MEMS based acoustic devices. Rock had already set up a number of other enterprises and is well known for starting companies and selling them at a tidy profit. He delights in this sort of activity and is a consummate presenter of information. At the global press event in Monterey where they unveiled the first offering, Rock's presentation was voted the best by more than 50 editors present. A major asset to the company is the legacy of Dr. Gabriel's previous work. The company has an exclusive license for CMOS MEMS from CMU for acoustic, inertial, and RF fields of use.
Akustica's first Microphone Chip is the first CMOS MEMS chip in the industry. It is a surface mountable, automatic pick-and-place compatible, monolithic device that provides high-quality voice input for consumer electronics applications. Akustica's patented CMOS MEMS technology is an industry first in both the new products it enables and the fabrication methods by which those products are manufactured. Unlike other MEMS devices which must be fabricated by special MEMS foundries, Akustica's CMOS MEMS structures are composed of the metaldielectric structures within a standard CMOS wafer. Since they are fabricated using the industry-standard CMOS processes and equipment currently used to make integrated circuits, CMOS MEMS chips can be manufactured in any CMOS foundry worldwide. In fact, CMOS MEMS device fabrication has been proven in nine different foundries and eleven different CMOS technologies, ranging from a 0.6 µm three-metal process to a 0.18 µm copper interconnect process. The result is a technology that can be mass-produced in extremely high volumes with the accompanying high yields and repeatability associated with semiconductor manufacturing. This makes Akustica's CMOS MEMS extremely attractive from both a performance and a manufacturing point of view. Without a doubt it is the ability of Akustica to manufacture their devices on existing CMOS platforms at high volumes that make their products so appealing. MEMS has always been slowed by the fact that each device requires a new tool set making them expensive to manufacture at first run. Akustica are already manufacturing their product in volume at foundries like TSMC.
The Akustica microphone chips are small, thin, light devices designed to replace the Electret Condenser Microphone (ECM), a fifty-year old technology that has been used in billions of portable electronic devices-while remaining fundamentally unchanged. The ECM, however, is limited by a number of acoustic issues: From RF noise, environmental noise, and mechanical/electromagnetic noise, the ECM cannot be used to easily eliminate all this mechanical and ambient noise-creating a number of pain points for millions of users of mobile phones, laptop computers and other digital media devices. As a silicon microphone, Akustica's Microphone Chips are not prone to the same degree of noise from any of these sources.
One of the fascinating parts of the Akustica demonstration of the new MEMS microphone was the initial use in laptops. The microphone in a lap top is the only part that cannot be implemented automatically as the temperatures involved in solder reflow will destroy the device. When the Akustica team first went to sell the product they approached laptop manufacturers and asked a question so basic that you wonder why nobody had ever reacted before. Why were the microphones always placed directly above the mechanical and noisy part of the laptop? Even without an improved device, positioning the microphone in the lid of the laptop automatically improves the position and functioning of any microphone. Laptop manufacturers were convinced and the first Akustica microphones will appear in the upper edge of a laptop lid directed to the user.
In the demonstration we were shown how the microphone chips suppress background noise and are selectively directional. Speaker chips will reproduce sound with greater fidelity and, with the added ability to detect motion, pressure or proximity, future chips can enable a wide range of end products to interact more intuitively with users. Sensory Silicon systems-on-a-chip are smaller, thinner, surface mountable, and fully compatible with automated assembly systems. Manufacturers will be able to design smaller, lighter, more power efficient products, and produce them more cost effectively.
With such groundbreaking news of a potentially disruptive manufacturing process I decided to put questions to Akustica directly to explain the technology. The patents involve manufacturing in the metal layer and when asked, the people at Akustica made it clear they have no intention of licensing the process and intend to release a new product every six months or so.
1.What have been the industry challenges that have required this technology?
There is a problem with current microphones:
Most microphones in consumer electronics (CE) products today are based on technology that has remained fundamentally unchanged for 50 years. Problems with the Electret Condenser Microphone (ECM) include noise, size, and manual assembly.
However, the main challenge for the audio system designer is to achieve the lowest overall noise in the system design. The noise of an ECM is a function of several sources: electrical noise resulting from fluctuations in the bias voltage, noise of the FET, board noise, acoustic self noise of the diaphragm, and external Electromagnetic (EM) and Radio Frequency (RF) fields that are coupled into the high impedance input of the FET.
2. What solutions have Akustica developed to meet industry challenges?
Single-chip digital, silicon microphone solutions developed on the CMOS MEMS platform free consumer electronic device designers and manufacturers from many of the problems associated with ECMs. When electronic circuitry is fabricated within microns of the acoustic structure, the short trace lengths lead to an inherently improved ability to mitigate RF noise. The CMOS MEMS microphone has a very short diaphragm to preamp distance and better input to output isolation due to the on-chip amplification stage as opposed to the FET in an ECM. Since there is better power supply and output signal isolation as well as a shorter distance between the diaphragm and the preamplifier, there is less chance of coupling EM fields into the microphone.
The AKU2000 CMOS MEMS microphones also integrate an analogue-to-digital converter on the chip, creating a microphone with a robust digital output. Since the majority of portable applications will ultimately convert the analogue output of the microphone to a digital signal for processing, the system architecture can be made completely digital, removing noise-prone analogue signals from the circuit board and simplifying the overall design. CMOS MEMS microphones also solve many of the mechanical design and manufacturing challenges associated with using an ECM. Firstly, the monolithic nature of the CMOS MEMS microphone enables a footprint and height that can be less than half that of a traditional ECM size. Secondly, the small size and mass of the CMOS MEMS microphone diaphragm which has a diameter of less than 0.5mm leads to improved vibration immunity as compared with an ECM which has a diaphragm diameter from 4-6mm. Thirdly, since CMOS MEMS microphones are fabricated using standard CMOS materials and processes, they are inherently able to withstand the high temperatures required for surface mounting. Therefore, no mechanical interconnect is required which leads to another significant reduction in overall height of the microphone system. Finally, the surface mount and pick and place compatibility of the CMOS silicon microphone reduces cost by eliminating manual assembly, thereby improving reliability, manufacturing throughput, and yield.
The widespread availability of well-maintained CMOS models and simulation tools results in products that can go from design to prototype in a matter of weeks. Leveraging the economies of scale, high quality, and maturity of the semiconductor industry, CMOS MEMS provides cost effective solutions that can be incorporated into mobile phones, digital devices, and automotive accessories.
3. How does Akustica's methods benefit the MEMS industry?
The technology represents a breakthrough in monolithic devices since, unlike other MEMS technologies, the structures are not fabricated in thin films on top of CMOS, but instead are fabricated from the metal-dielectric layers of the CMOS itself that are deposited during the standard CMOS processing flow.
As the first MEMS devices manufactured using standard CMOS processes, Akustica's Microphone Chips can be manufactured in quantity, with guaranteed uniformity, by existing CMOS foundries-rather than by dedicated MEMS foundries. Easier manufacturing increases access to the chips, lowering costs. CMOS MEMS chips soon will become as commonplace as standard CMOS devices. Thanks to widespread CMOS models and simulation tools, they certainly will be as familiar and as easy to use. Leveraging economies of scale, high quality, and maturity of the semiconductor industry, CMOS MEMS will offer a cost-effective solution that goes from design to prototype in a matter of weeks.
4. Explain the potential of the microphone and the practical implementation (including position in laptop)
An important feature of today's mobile computers is the ability to provide high quality voice input for hands-free communication. The optimal solution to support this application is to embed the microphones directly into the bezel of the laptop display. While an embedded microphone array is ideal from a performance point of view, the laptop designer is faced with mechanical, electrical, and acoustical challenges when developing an integrated platform. The AKU2000 digital-output microphone has been designed to overcome these challenges and enable simple integration of a microphone array into a laptop PC.
The AKU2000 exhibits many qualities that make it ideal for integrated microphone array applications in laptop and desktop computers. Most importantly, the robust digital output is immune to the EM or RF interference that can prohibit optimal acoustic placement of a standard analogue-output microphone in a laptop computer. The small footprint and thinness also increase the flexibility of the microphone placement.
5. What are the future potentials for this technology?
Microphones are only one possibility. CMOS MEMS will enable single-chip integration of all kinds of electro-mechanical sensor structures with analogue and digital signal processing functions to produce acoustic, inertial, and RF systems-onchip.
As a result, countless electronic products will be able to hear, speak, and sense the world around them.
