Xun Liang | PubMed Health

PubMed2005

A novel microcomputer temperature-compensating method for an overtone crystal oscillator.

In this paper, a novel microcomputer temperature-compensating method for an overtone crystal oscillator (MCOXO) is presented. In this method, a ceramic oscillator is chosen, and its output frequency is mixed with the output frequency of an overtone crystal oscillator. A crystal filter is used to suppress the spurious mixing products. A microcomputer is used to control the switch capacitance array that is connected to the ceramic oscillator circuit. The frequency deviation of the crystal oscillator is directly compensated by the output frequency of the ceramic oscillator. As a result, the method is able to overcome the disadvantages of frequency stability degradation and phase noise deterioration that are provoked by adding inductance or frequency multiplication in traditional compensating approaches. At the same time, this method is able to compensate a quite wide frequency range and many types of oscillators, not just crystal oscillators. The experimental compensating results show that, using this method, the frequency-temperature stability of a 100 MHz 5th overtone temperature-compensated crystal oscillator can achieve < or = +/-2 x 10(-6) for 0-70 degrees C.

Journal Article↗

PubMed2004

Complexity of error hypersurfaces in multilayer perceptrons.

Error hypersurfaces are very valuable to study because of their unique status in multilayer perceptron research. Given the architecture of a multilayer perceptron, if the pattern sets are different, so are the respective error hypersurfaces in the multilayer perceptron. Using the theory of groups and Polya Theorem, this paper constructs classes of congruent pattern sets and classes of congruent error hypersurfaces, and proves that the number of classes of congruent pattern sets is equal to the number of congruent error hypersurfaces. Calculation results lead to much fewer classes of congruent error hypersurfaces than the total error hypersurfaces, and show that as the input dimension N increases, the former number increases at a much lower rate than the latter number, thus simplifying the understanding of the complexity of classes of error hypersurfaces.

Algorithms↗