The subject of this thesis is the investigation of a digital absolute posi- tion-detection system, which is based on a position-information carrier (i.e. a magnetic tape) with one single code track on the one hand, and an array of magnetoresistive sensors for the detection of the information on the other. The sensor array consists of a permalloy (Ni0.81Fe0.19) strip which is electrically sub-divided into adjacent sensor elements by means of cross- contact leads (Mo-Al). The sensor response on a magnetic field in the plane of the strip is a result of the anisotropic magnetoresistance effect in thin ferromagnetic films. The permalloy strip has a thickness of about 50 nm and a width in the range of 10 to 30 um. The central positions of the sensor elements are located at mutual distances of 50 to 200 um, which is dependent on the bit period in the magnetic tape. The plane of the sensor strip is positioned perpendicular to the magnetic-tape plane. The strip axis and direction of movement of the permalloy strip are lengthways along the magnetic tape. This tape is recorded with a so-called pseudo-random bit pattern, having the property that each arbitrary selected set of n adjacent bits is unique. Depending on the actual position of the sensor head, a special bit combination is detected by a sensor array con- sisting of n elements. This digital word can be decoded to a position coordi- nate. The thesis reports on the feasibility of the position-detection system on the basis of experiments with specially developed measurement heads. In introduc- tory observations the backgrounds, the theory and the design choices come up for discussion. Attention is paid to the modelling of the magnetization distribution in the permalloy strip, which is excited by homogeneous or inhomogeneous magnetic fields. Next the conversion of the magnetization distribution into magnetoresistance and so-called planar-Hall voltages is treated, resulting in a clear choice for the application of the former in a position-detector array. The potentialities of a silicon Hall-sensor array as an alternative for permalloy sensors is considered. Next the basic set-up of a complete position-detection system is discussed, including the generation of the applied position code, as well as methods to cope with problems like the sensor arrays being non-ideally positioned with respect to the bit centres and the potentialities of the application of interpolation schemes. For practical reasons commercially available, longitudinally magnetized tapes are used as a position-data carrier in an experimental system. In view of the computations of the sensor behaviour in the stray field of such a tape, the foundations of a computer simulation model are discussed, resulting in both a practical "simple" model and a more-difficult sophisticated model. Next the details of the realized sensor-chip design are outlined. The realization of the sensor head has been a very demanding activity of this research project; two subjects are discussed in detail. First the realization of the required low interface-resistivity value between permalloy and conduc- tor film is discussed. Second the positioning of the sensor strip at the very edge of the substrate (distance in the order of micrometers) is described. The measurements chapter describes the applied measurement equipment, the various properties and dimensions of the prototype sensor structures and the set-up of the actually performed measurements, i.e. the registration of the sensor-element responses in the stray fields of magnetic tapes. The results are evaluated for sensor elements and tapes with various properties, in the light of the so-called detection distances which are determined with the help of "eye patterns" composed on the basis of worst-case responses. The lower limit of the bit period is found to be in the order of a few hundred micro- meters. Conclusions are first addressed to the performance of the computer model. It is found that the simple simulation model results in reasonably accurate approximations of the sensor responses and detection distances, if the computations are carried out for relatively small values (t ~ 10 um) of the separation between the sensor strip and the magnetic-tape surface. A careful extension of the range of the experimental parameters in the computer model does not result in essentially better computed figures for the minimum bit period (i.e. resolution) of the position-detection system: Prospects of improved performances could possibly arise by the introduction of an all- embracing second design round. It is noted that the final resolution of the position-detection system, being in the order of magnitude of 100 um, can be improved by a factor of 10 to 100 with the application of interpolation techniques. In this case the increase of the accuracy which, in the first instance, is in the order of magnitude of the bit period, will be dependent on the actual quality of these techniques. Finally, the performance of a position detection system based on the alternative silicon-Hall technology is estimated to be of the same order as the permalloy-based version. In an appendix attention is paid to measurement techniques which, based on magnetoresistance sensors, at the time were applied for analysis of the form and magnitude of the compact head field of recording write heads. The utilization of the discussed absolute position-detection system should be based on the possibilities to economically realize a (mass)product with relatively good performance with respect to the established (mostly optical) systems. Having this aim in view, a special-purpose signal-processing chip should be developed, among other things, and integrated with the thin-film sensor structure. However, extensive and multidisciplinar activities of this kind are found to form a certain barrier for the utilization of the design.
|Qualification||Doctor of Philosophy|
|Award date||1 Oct 1990|
|Place of Publication||Enschede|
|Publication status||Published - Oct 1990|
- SMI-REC: RECORDING
- SMI-TST: From 2006 in EWI-TST