Current Projects
Note: several of these projects come under the heading of LARSS, for Large
Area Restructurable Silicon Systems. It is a variation on the Restructurable
VLSI areas that I have used to create wafer scale integrated circuits during
ten years of research at Massachusetts Institute of Technology's Lincoln Lab.
It was briefly described above but best described in the papers. I have
several projects in the design of LARSS/wafer scale designs, ranging from some
specific circuits to the more generalized problems.
The first is the creation of Large Area Restructurable Silicon System (LARSS) which involve arrays of both integrated circuits and transducers. This work aims at the creation of new silicon systems that can readily be expanded to areas and complexities much larger than current integrated circuit fabrication limits. To increase circuit area to the wafer scale level, 10-200 times the current IC area, requires defect avoidance technologies bypass IC fabrication errors and hook together working sections to create large systems. The technique divides the system into redundant circuit blocks surrounded by a grid of metal signal/clock/power bus lines to which the devices can be connected. In all cases the designs use standard IC fabrication services (eg. Canadian Microelectronics Corp.) to manufacture the devices, but applies post fabrication processing to achieve the unique systems. One interconnecting technique which makes permanent signal routing uses a laser linking system developed by the applicant at M.I.T. Lincoln Laboratory and currently applied at Simon Fraser University. It consists of two separated dopant lines in the silicon substrate which are connected to the bus lines (like a gateless transistor). An external laser melts the gap area causing the dopant lines to connect with a resistance of typically 60 ohms. The laser can also cut metal lines, thus segmenting the signal buses.
In terms of work here my students and I have built a major unique piece of experimental equipment. This is a laser interconnection table which will focus a controlled laser beam on integrated circuits with a 0.1 micron position repeatability. This is one of only five such systems in the world, two at M.I.T. Lincoln Lab, plus one each at the University of South Florida and the US National Security Agency. I have played a major part in the creation of all such systems, and feel this is the best designed so far. The value of this equipment is demonstrated by the interest in using the laser table shown in the Industrial Cooperation section.
Due to the expensive nature of building true wafer scale systems my work has concentrated on studying techniques for building large area systems and the design of new classes wafer scale type devices. At the design level this has resulted in building circuits to compare the area and speed tradeoffs for defect avoidance of laser connections and active devices, which has lead to the design of circuits combining both for optimum operation. My research has proposed and tested a new type of large area product: the Wafer Scale Field Programmable Gate Array which could increase the density of current FPGA systems by 10-100 times. Another new concept was LARSS systems containing both transducers (some micromachined), control circuitry and the laser redundancy network needed to build in large areas. First proposed in one of my papers, the idea here is to build large transducer arrays using separate types of redundancy layouts for both the transducers and the circuitry. One such systems is a large area magnetic field sensor array designed for mapping magnetic field distributions being undertaken by my PhD student Yves Audet. Another array was a design for a wafer scale array of thermal emission elements. In this transducer matrixes the laser redundancy scheme changes the design from something that can be used only in small chip sized structures (say 10x10 arrays) to very large devices possibly of 256 x 256, which makes it of considerable interest to several organizations.
The area of true micromachining and the development of microfabrication techniques is of considerable interest to me. As noted above micromachining and transducers is playing a significant role in the many of the large area systems that are currently the research of myself and my graduate students. In addition to classic etching micromachining I have in particular been looking at Laser Micromachining which uses laser beams to create structures in the 50 micron and below size. My micromachining developments has concentrated on three areas. First was the development a new micromachined vacuum sensor that has been the subject of a MSc and BSc thesis by students, and is expandable into creating very sensitive flow sensors. Two of my other sensor developments were created as part of my Industrial Cooperation. The first is a laser micromachined vision/proximity sensor called "Vision Skin" for the Canadian Space Agency, which both gives a low resolution image of the an approaching object and measures the distance to it with precision 100 micron precision. The second involves developing advanced X-ray detectors with micromachined sensor electrodes. Included in this is several new characterization techniques we have developed.
Several projects connected to embedded microprocessor systems are currently being started with local industry. These include computer antithief devices and advanced development software.
In the publication area I, together with my students, have been published s total of 26 journal papers, 47 referred conference papers, and two book chapters published. I also have 7 patents issued, and two patents applied for. One Patent "Method and Apparatus for Forming Low Resistance Lateral Links in a Semiconductor Device" is currently under licensed by a company. Another two patent applications being payed for by the Canadian Space Agency, and they have licensed the work to Kinetic Sciences Inc.