Analysis of fault characteristics and control of microelectromechanical systems (MEMS) is becoming increasingly important as these devices are incorporated into safety critical systems. Mechanically decoupled monitoring techniques are needed to ensure that accurate, low-noise positional information is gathered from these devices during operation. Optical monitoring has been shown by various groups to be capable of achieving this goal. However, Integrated Optical Monitoring (IOM) must be achieved in order to enable incorporation with the MEMS die in a single package for effective lifetime system monitoring. This research addresses the design, fabrication and testing of diffractive optical elements (DOEs) for achieving IOM.. DOEs for on and off-axis single and multiple beam generation are modeled based on a linear optical system model paraxial gaussian beam theory. The DOEs are generated using a MATLAB tool that was completed as part of this work to calculate the desired input wave, output wave, and transfer function. This data is then used to generate a phase lens profile with an etch depth that is dependent on the material that will be used for lens fabrication. DOEs are fabricated on a borosilicate substrate and testing of the lenses is performed using a beam profiler to determine the number, shape and power of the probe beams for comparison with the results from the simulated lenses. Experimental DOE test results compare well with those obtained from the modeling tool validating the use of the DOE design tool and the fabrication process for subsequent research implementing IOM architectures.

No abstract

No abstract

This article documents investigations into impedance control carried out on a direct drive robot manipulator. Both linear and nonlinear versions of impedance contrallers have been investigated. The two impedance controllers have been simu lated numerically and implemented on Carleton University's direct drive robot. The performance of these controllers is ex perimentally compared in terms of force trajectory following and contact stability. The article highlights the practical and experimental limitations of the proposed methods. Furthermore, the experimental results indicate that the controller design, the manipulator kinematics, and the desired task must all be considered and evaluated simultaneously.

No abstract