![]() A display loop has been implemented to provide visual feedback at a refresh rate of 20 Hz. Indeed, the mechanical bandwidth of human movement is around 7 Hz (2 Hz for normal speed movements). This frequency is sufficient for control because friction in the system provides stability and because human motion is characterized by a low bandwidth. The main control program analyzes position and force inputs and calculates commands to send to the motor at 100 Hz. Data from force sensors are sampled at 1000 Hz by a data acquisition card (USB-6211, National Instruments). Data from the EPOS controllers of the motor (positions, velocities and currents) are read at a frequency of 100 Hz and transferred to the main program using an RS232 protocol. The main program is divided into subtasks to separate control, display, and data acquisition, and thus distributes the tasks and allows faster control. The interface is controlled by a program written in LabView 8.2 (National Instruments) that runs on a PC (Pentium 4,4 GB RAM, 233 MHz). The gear ratio between the motor shaft and the clutch is 2. The interface is actuated by one brushed dc motor (Maxon motor, Switzerland RE40, 150 W encoder 500 counts/rev control card EPOS 24/5). Plastic cogwheels wore out quickly, therefore, steel cogwheels were more suitable for our purpose. Cogwheels, which are used for the clutch system, made from POM or steel were tested for durability and transmission smoothness. ![]() The 30 pulleys used to guide the cables are made of POM (polyoxymethylene, or Delrin) and are mounted on standard ball bearings. Polyester, polyester reinforced with carbon fibers, polyethylene fiber, and steel wire were tested and it was found that steel cable (diameter of 0.5 mm) had the most suitable combination of these factors. Materials for the cable were compared according to their compliance, friction around a pulley, breaking strength and creep. Insertion of just the fingertip was mostly used for the feasibility study with stroke patients. The finger can be inserted partially or completely into the Velcro loop, i.e., just the fingertip (distal phalange), or as far as the intermediate or the proximal phalange. A better finger fixture consisted of a Velcro loop within a metal ring to which the two ends of the cable were attached (Fig. The size of the thimbles was not adjustable and subjects reported discomfort due to perspiration after using them for more than 20 min. A first method was tested where the subjects inserted their fingers into sewing thimbles. The use of gloves is precluded in our application as stroke patients with spasticity may have difficulty in donning them. Different techniques for finger attachment are possible. The support can be adjusted to change the position ( 10 cm) and the orientation (☓0°) of the forearm (Fig. The support for the forearm and elbow was designed to provide comfort, while mechanically isolating the hand from other body movements. This sensing system is suitable for our purpose as only forces parallel to the cable are measured. The variation is linear within this interval, i.e., the correlation coefficient with the linear fit is 0.96. Therefore, the workspace is constrained to a central interval, i.e., 3.5 < a < 21.5 cm, where the variation of the force represents less than 3% of the nominal force. The force is significantly biased when the force F is applied near one of the pulleys P 1 and P 3. The coefficient k of 40 N/mm is the compliance of the system.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |