R.B. Yates

A design methodology for linear micro-generators is developed, and is applied to the design of a mm scale electromagnetic micro-generator. The fabrication of a prototype device is also described using generally available microfabrication techniques, and the results of testing the device on a variable amplitude vibration source, in air and vacuum, are presented. The experimental results confirm the design rules and indicate how the generation of useful power levels might be achieved.

Supplying power to remote microsystems that have no physical connection to the outside world is difficult, and using batteries is not always appropriate. A solution is offered by the device proposed in this paper, which generates electricity from mechanical energy when embedded in a vibrating medium. This micro-generator has dimensions of around 5x5x1mm. Analysis predicts that the power produced is proportional to the cube of the frequency of vibration, and that to maximise power generation the mass deflection should be as large as possible. Power generation of 1/spl mu/W at 70Hz and 0.1mW at 330Hz are predicted for a typical device, assuming a deflection of 50/spl mu/m.

Orthograde axonal transport tracing methods were used to describe the projections to the basilar pontine nuclei (BPN) which arise within the face representation of motor or somatosensory cerebral cortex. Injections centered in motor face (MF) cortex resulted in the labeling of several corticopontine terminal fields which exhibit a rostrocaudal columnar arrangement within the ipsilateral BPN. The location of such terminal zones is consistent with the somatotopic pattern of termination previously described for limb sensorimotor cortices. In contrast, the projections from somatosensory face (SF) cortical regions largely terminate in BPN areas separate from those receiving either limb sensorimotor or MF inputs. Both MF and SF cortices also give rise to projections to the contralateral BPN; those from SF cortex are less extensive than those of MF origin. In addition to their relationship with limb sensorimotor corticopontine terminations, the MF projections to the BPN also seem to partially overlap the projection zones of the cerebellopontine system, particularly the regions projected upon by the lateral cerebellar nucleus. The SF projections, on the other hand, appear to terminate in BPN areas that also receive input from either the dorsal column nuclei or the spinal trigeminal complex. There is only minimal potential overlap between MF and SF projections in the BPN. With regard to the pontocerebellar system, the projections from MF cortex terminate among BPN neurons which project to the cerebellar hemispheres, particularly lobus simplex, crus I and crus II. The SF projections also overlap BPN neurons which project to the lateral hemispheres in addition to the paraflocculus and vermal lobules VII and IXa,b. Taken together these observations suggest that subsets of BPN neurons might exist such that some receive convergent inputs from systems whose function can generally be regarded as motor (sensorimotor cortex, cerebellopontine) while another population of BPN neurons might integrate signals from systems which transmit somatosensory information (dorsal column nuclei, spinal trigeminal).