Abstract: An exercise machine has a rigid base on which stands a person wishing to exercise. One or more handles extend from the base and are each connected thereto by a friction pivot assembly. The person exercises by moving the handles against the friction of the pivot assemblies. Each pivot assembly has two or more physically fixed non-parallel axes about which the corresponding handle rotates, and this allows a wide range of movement of the handles. The pivot assemblies have bearings for providing frictional resistances to movement about each axis, and those resistances are independently adjustable.

Abstract: A position sensor has a detector having a plurality of sensitive elements which are acted on by indicia of a scale so that each sensitive element detects overlap between itself and one or more indicia of the scale. The outputs of the sensitive elements are modified by weighting values so that there is one position of detector and scale where the sum of the outputs has a unique characteristic. This position can be detected by means analyzing the sum of the outputs, and can then detect displacement of scale and detector from that one position. The indicia used have a non-uniform pattern, but other indicia with a pitch less than twice but not equal to the pitch of the sensitive elements may be provided to permit interpolation to fractions of the pitch of the sensitive elements. Further interpolation may be provided by a register through which signals propagate at a rate which is a multiple of the rate they are received.

Abstract: A position sensor has a scale consisting of tracks of indicia in the form of bar patterns acting on a detector having a two-dimensional array of sensitive elements. One track has indicia with a pitch slightly greater than the pitch of the sensitive elements and another track has indicia with a pitch slightly less than the pitch of the sensitive elements. The other tracks provide a coarse determination of the position of the scale relative to the detector while the first two tracks provide resolution of the position to less than the pitch of the sensitive elements. There is sufficient redundant information to enable accurate position detection even if there are optical or electrical perturbations in the detection of individual indicia.

Abstract: In order to provide a laser beam having a relatively constant width (i.e. transverse spacing between points receiving a theshold irradiance), a converging lens 15 having substantial negative spherical aberration is used. In the far field, the beam comprises paraxial rays 18 which have diverged to provide the desired beam-width. In the near-field, the smaller, higher-irradiance beam formed by these rays is augmented by a `sheath` of marginal rays 19 which are not refracted towards the optical axis as strongly as the paraxial rays. The shape of the beam in the far field is determined primarily by the shape of the (laser) light source 10, while that of the near-field beam is controlled by a mask 14. In another arrangement, a lens 102 having substantial positive spherical aberration is used to create a `light source` having an accurately-controlled brightness distribution and very small size.