Die implementering van ’n kleinste-kwadrate- algoritme vir die frekwensiebepaling van ’n protonpresessiesein

An advantageous signal processing technique for estimating the frequency of a time-limited noisy sinusoidal signal has been implemented. The method is a least squares technique and is especially powerful when the estimation has to be done in the shortest possible time measured in terms of signal period, as is the case with the precession signal present in a proton magnetometer. The criterion for evaluating the performance is the minimum measuring time for a given error or the minimum error for a given measuring time. This technique is compared with the accepted way of measuring and the results show a considerable improve­ment