The IMPACT team sought to discover whether analogue and radio frequency (RF) circuits could be developed for the emerging 90nm CMOS chips. These chips use a much smaller (90nm) circuit etching process than current models.
"The 90nm digital CMOS chips are becoming available this year and we wanted to know if we could use them for high-frequency microwave communications," says Dr Stefaan Decoutere, IMPACT project coordinator and researcher at IMEC in Belgium.
Older RF-CMOS chips are already used in 2.4Ghz devices like the Bluetooth transmitters and receivers developed separately by Ericsson, one of the project partners. Currently these applications require expensive, multi-chip solutions.
IMPACT found that the 90nm CMOS chips can meet the stringent specification standards for mid-range and high-end applications like GSM and Point-to-Point microwave communications. Ultimately it will cut costs, reduce power consumption and increase the functionality of these devices.
"These chips are expensive to develop but over time they are much cheaper to produce," says Decoutere. Because all the circuits are on one chip the production process is much more efficient and RF-CMOS chips can be produced in large volumes, slashing costs.
CMOS chips are an important type of integrated circuits. They include microprocessors, microcontrollers, static RAM, and other digital circuits. These different functions can be integrated on a single chip, further reducing costs and the complexity of devices like digital cameras. CMOS chips only use power when its transistors are switching between on and off states. As a result, they use less electricity and generate less heat than rival chips, which means they are ideal for mobile applications.
The IMPACT team scored two successes. Firstly, they achieved a given performance specification at significantly lower cost and power consumption than can be managed with today's technologies. Secondly, they pushed the state-of-the-art circuit performance in RF-CMOS technology with demonstrator low-noise amplifiers and voltage-controlled oscillators with world-record figures-of-merit in performance and robustness against Electrostatic Discharge.
The project won the enthusiastic support of the industrial partners. Applications for the 90nm CMOS technology are already in the works. For example, Ericsson's target applications over the next three years include WLAN systems, like WiFi, and terminals in the 5 to 6 GHz frequency range.
Project partner Philips says that the use of mainstream CMOS technology as a cost-effective solution for wireless communication is a particularly important emerging market with a huge range of potential product applications. Philips believes this technology will become an enabler for wireless consumer electronics and that virtually all-electronic products may become wireless so as to enable device-to-device interaction. This not only includes high-end and luxury products, as are available now, but also but also low-end products such as Walkmans and MP3 players.
But these devices are heavily dependent on the timely integration of RF and analogue functionality into advanced CMOS technology to yield cost-effective system-on-a-chip solutions. IMPACT has gone a long way to solving these problems.
In the meantime, IMPACT wants to launch a follow-up project to see if it is possible to develop CMOS chips using a 45nm etching process. Advantages in a 45nm CMOS chip could include smaller chip size, the potential for improved performance, and lower power consumption for the same functionality, if the researchers can get it to work.