To drive a step-motor, the open-loop method is commonly used. In this method, the motor makes one step for every one pulse, accordingly the highest repetition rate of applicable command pulses is limited by the time of initial one step of the motor from its standstill state.
This paper presents a closed-loop control method, setting a position-coder to a step-motor shaft. In this latter method, the motor makes self-commutation ranning using feedback pulses, when command pulses with high repetition rate are suddenly applyed. Using a reversible counter, misssteps are avoided by comparing feedback pulses with command pulses. And a sudden stop of a rotor is possible by the proper choise of a exciting field, when contents of the reversible counter become zero.
By the application of this new method, it is possible for a given step-motor to responds to much higher rate of command pulses than the case of the open-loop method.

View full abstract

For the analysis of non-linear control systems, the phase-plane method and describing function method are currently used. The former is exact but very tedious way if the system is not very simple ; the latter is easy to use even for the complicated systems, but it suffers an appereciable error for the system with strong non-linearity.
In order to make the philosophy of the describing function applicable even to the system with strong non-linearity, a set of functions, temporarily named as saturation functions, is first introduced, where the square wave, triangle wave and so forth are represented by the saturation functions of the zero-order, first-order and so forth. Then several fundamental relations between these functions are explained, and these are compared with the counterparts of the trigonometric functions.
The method is applied to the analysis of certain types of control systems with strong non-linearity, and it is found that the describing function method is incorrect while this gives an exact result.

View full abstract

Monte-Carlo methods have been used as a method of numerical calculations, but these applications are limited to large scale digital computers.
However, if we prepare random voltage source, as discribed in this paper, we can do integral operations of analog voltage comparatively easily through Monte-Carlo methods, and we call it as “Random voltage methods”.
In this paper the author presents a prototype of random voltage source using thyratron noise, flip-flop and D-A conversion circuits, and also shows the basic principle of the integrator of the analog voltage waveforms. As an example, the value of π has been calculated by this integrator.

View full abstract