A design method is described for a new analog-to-digital converter (ADC) which can vary its effective resolution depending on the input frequency. This is an important development, especially for video signal processing. Usually, 8 bits are sufficient for a video signal, however, when the video signal frequency is very low, 10-to 12-bit quantization becomes necessary. The ADC is based on a subranging ADC, which consists of coarse and fine converters. The ADC can be contrasted with the conventional subranging ADC in which the fine conversion cycle follows the coarse conversion cycle. A prediction method is adopted to make both the fine and coarse conversions at the same time. The lower the input frequency is, the more accurate the prediction is. As a result, the effective resolution of the ADC depends on the input frequency. To estimate the ADC performance, an experimental circuit has been implemented with discrete circuit components. The effective resolution change with change of input frequency was measured. Experimental results were in good agreement with the theory.
This paper shows the experimental results of the flux control characteristic and inductance controm characteristic of magnetic fluid which is controlled by the permanent magnets. Two kinds of magnetic field, parallel and orthogonal, are comparatively discussed. The position between the magnetic fluid and the permanent magnet determines the magnetic field. The results are useful in using magnetic fluid as a sensor or a control element.
Numerical simulation and the corresponding experiments have been conducted to apply isothermal capacitance transient spectroscopy (ICTS) technique to the study of Si MOS diode. The numerical simulation for the ICTS spectrum of MOS diode has been performed in consideration of both the interface states and bulk traps. The numerical simulation shows the detailed behavior of the ICTS spectrum on the interface states. As a result, it is ascertained that the spectrum of the interface states can be identified using a reverse bias technique where dependence of the spectrum on a reverse bias voltage is investigated. Using gold-diffused MOS diodes, an enhancement of the interface states by the gold diffusion has been observed. Although the gold diffusion process is inevitably accompanied with thermal annealing, the influence of the thermal annealing on the interface states is eliminated with the use of an extra annealing at a low temperature in a nitrogen atmosphere. It is also ascertained that the interface states enhanced by the gold diffusion depends on the density of diffused gold.
One of the common problems in the neuro-computer hardwares is complication of wiring. Folthret as a basic element for the neuro-computer, which has been offered by one of the authors of this paper, can solve this type of problem quite well. Therefore this element is advantageous to make up the neuro-computer. In this paper, Folthret has been at first realized with the hybrid style of analog and digital circuits for the purpose of investigating its size. Resultantly it is composed of a printed wiring board with the size of 15cm×15cm, and 26 integrated circuits. Half the printed wiring board is occupied by the memory unit to retain the weights, because the memory unit is composed of digital memories, A/D and D/A converters. Thus Folthret will become more compact when an appropriate CCD with large capacity is developed. Secondly with the Folthret realized by electric circuits, a learning experiment of various kinds of two-class pattern classifications has been conducted by using 20 random patterns for the purpose of evaluating its learning capability. As a result, it was revealed that Folthret can learn every two-class pattern classification mentioned above. Also the number of learning cycles etc. have been confirmed to be fairly in accordance with the case of the simulation using of a learning threshold-element model.
In conventional bill money recognition machines, we develop the recognition algorithm according to the transaction speed and diffierence of various specifications. However, development of the algorithm for the recognition has been based on the trial and error method. Many researchers have reported that neural networks are suitable for pattern recognition because of the ability of selforganization, parallel processing, and generalization. In this paper, we present a new bill money recognition method with neural network and show the effectiveness of the present algorithm compared with the conventional method by discrimination functions. Furthermore, we transform bill money data by FFT into frequency domain to reduce influence of the noise due to conveyed fluctuation. Then we adopt these Fourier coefficients or its amplitudes as input of the neural network. We show that the ability of recognition can be evaluated in detail by introducing a new measure of reliability.
Pb battery has become to be used for the power source of electric vehicle, because of its long life and low cost among the other kinds of batteries. In this case, the range of the vehicle is not so long and the number of charging stations is limited. Thus, in order to predict the residual range, it is necessary to estimate the dischargable value for Pb battery. The open voltage of Pb battery can be used to obtain the above estimated value. Generally, to measure the precise open voltage, it takes much time after cutting the circuit to interrupt discharge. In this paper, we first explain the experimental system which simulates the electric vehicle and Pb battery discharge. Then, we propose the prediction method of the open voltage by adaptive digital filter (ADF), which enables us to get the predicted value continuously while driving electric vehicle. Based on the predicted value, we propose the method to estimate the residual Wh values. Finally, we apply the present method for experimental real data to show the effectiveness of the approach.