The amplification of a 10 ps pulse by two-stage electron-beam pumped KrF laser amplifiers is reported. The energy of 4 J was obtained for the pulse, which is the largest energy ever reported for short pulse excimer lasers. Nearly 90% of the final amplifier (29cm diameter) was filled by the beam, and the energy density of the picosecond beam reached 3.9times the saturation energy density. 56% of the stored energy was estimated to be extracted to the pulse. Measured energy of the amplified spontaneous emission (ASE) agreed well with the value estimated theoretically, and most of the ASE was observed to originate in the first amplifier. From the comparison of the measured energy, pulse duration, and divergence of the picosecond pulse to those of the ASE, the S/N ratio of the picosecond pulse to ASE is estimated to be as large as 105 in terms of the focused power density.
We have developed a 100 W copper vapor laser driven with an all solid state exciter. The exciter consists of static induction (SI) thyristors and two stage magnetic pulse compression circuits with saturable reactors. In order to design a high power and high efficient copper vapor laser with an all solid state exciter, we used the data measured with a 20 W all solid state copper vapor laser which was demonstrated for the first time previously. As the result, we achieved a laser output power of 103 W and a maximum overall switching efficiency of 84% at a 5 kHz repetition rate. Furthermore we succeeded in reaching down 4 ns of an overall system jitter for synchronization.
It is anticipated that the advent of fiber optic communications will lead to revolutionary concept changes in the field of communications in the coming fiber optic age. Signal multiplexing technology would by no means be exempt from this revolution. Existing multiplexers can only multiplex signals whose clock rate are close to integer multiples of each other. The wide band features of optical fiber can remove such restriction to attain great flexibility in multiplexing asynchronous signals. Here, we report a new multiplexer using previously reported adaptive run-length transform technology. This unit can adapt to differences in the multiplexing hierarchy. This experimental multiplexer has a maximum of 16 multiplexing channels. A transmission clock of 32 Mb/s and an information carrying capacity of 21 Mb/s. The experimental multiplexer provided errorless transmission. A new frequency locked loop is developed and used in receivers to avoid waveform distortions caused by asynchronous transmission. Jitter is found to be less than 10%, which is satisfactory for practical use.
As a zero order hold for a D/A conversion of a conventional digital time function generator is used to approximate a signal between two consecutive sampling instants, many sampling points are necessary to improve the approximation of the continuous signal. Consequently, the frequency range of the generated functions is restricted. For this problem the authors propose a new type of a time function generating method using delta modulation, which consists of a combination of the zero order hold of periodically sampled signals and the interpolation of the signal between its sampling points. In this method the signal between consecutive sampling points is interpolated by the sequences of the optimum positive and negative steps of the delta demodulator. Therefore, it is very important to decide the pulse height of the delta demodulator. The calculation of this pulse height is based on the spline function which estimates the signal between two sampling instants from five sampling data. The proposed function generating method has such features that the operational circuit of the optimum pulse height can be constructed by only analog adders, since it is caluculated a linear combination of the five sampled data, and that a good accurately approximation of a function and wide frequency range of the generated function can be obtained also.
The demand for electricity in the future is uncertain and difficult to forecast. But we have to forecast the demand in the future to construct power plants. In that sence a large scale power plant has a disadvantage, since it takes long to construct it. On the other hand, a small scale one is more expensive than a large scale one, although it takes less time to construct it. Therefore large and small scale power plants could complement the disadvantages of each other. In this paper we have developed a stochastic dynamic programming model to investigate the possibility. The model determines the capacities of large and small scale power plants in the uncertain demand to minimize the expected value of the total cost. Computed results have shown that small scale plants can play a significant role in the uncertain demand even if they are more expensive than large scale ones.