Robert M. Lofthus

Closed loop low-velocity regulation of hybrid stepping motors amidst torque disturbances

To regulate the velocity of hybrid stepper motor motion control systems, a control law which exploits the nonlinear dynamics to create an analog positional control in conjunction with a traditional linear control is introduced. This nonlinear approach allows a much coarser position sensor to be used, including position estimates based on back EMF measurements. The form of the control law admits the use of a wide variety of compensators, whereas earlier laws use only velocity damping compensation. Two specific compensators, i.e., velocity damping and integral control are analyzed in detail, then compared to each other and to open loop microstepping control. It is shown that velocity damping allows the design of the eigenvalues of the closed loop system and provides a linear system approach about a specified operating point. Unfortunately, this operating point includes the value of external DC torque (drag) present, so the closed loop dynamics cannot be guaranteed amidst steady state torque fluctuations. Integral feedback (within a PID controller) improves upon velocity damping by not only allowing the design of the closed loop eigenvalues, but also by completely linearizing the system regardless of external DC torque values. Furthermore, the integral feedback produces zero steady state position error (as expected from linear control theory) and significantly decreases the tendency of the motor to lose step. Experimental results validate the analyses. >

Hybrid integration of light-emitters and detectors with SOI-based micro-opto-electro-mechanical systems (MOEMS)

A multidisciplinary team of end users and suppliers has collaborated to develop a novel yet broadly enabling process for the design, fabrication and assembly of Micro-Opto- Electro-Mechanical Systems (MOEMS). A key goal is to overcome the shortcomings of the polysilicon layer used for fabricating optical components in a conventional surface micromachining process. These shortcomings include the controllability and uniformity of material stress that is a major cause of curvature and deformation in released microstructures. The approach taken by the consortium to overcome this issue is to use the single-crystal-silicon (SCS) device layer of a silicon-on-insulator (SOI) wafer for the primary structural layer. Since optical flatness and mechanical reliability are of utmost importance in the realization of such devices, the use of the silicon device layer is seen as an excellent choice for devices which rely on the optical integrity of the materials used in their construction. A three-layer polysilicon process consisting of two structural layers is integrated on top of the silicon device layer. This add-on process allows for the formation of sliders, hinges, torsional springs, comb drives and other actuating mechanisms for positioning and movement of the optical components. Flip-chip bonding techniques are also being developed for the hybrid integration of edge and surface emitting lasers on the front and back surfaces of the silicon wafer, adding to the functionality and broadly enabling nature of this process. In addition to process development, the MOEMS manufacturing Consortium is extending Micro-Electro-Mechanical Systems (MEMS) modeling and simulation design tools into the optical domain, and using the newly developed infrastructure for fabrication of prototype micro-optical systems in the areas of industrial automation, optical switching for telecommunications and laser printing.

Step motor supply: Minimizing torque ripple induced by digital linearization

Abstract A model of the torque ripple induced when linearizing hybrid stepping motors is developed. Using this model, the optimal waveforms for linearization are derived. These optimized waveforms require only a slightly different table look-up, yet they reduce the square root of the mean square torque ripple approximately five-fold, thus making square wave commutation possible for some applications. Frequency analysis of the resulting torque ripple provides a method of trading-off linearization computations for performance. Experimental results validate this analysis.