Bilateral sevomechanisms are generally loaded by quite different mechanical systems, where the case by case construction of transfer function is inconvenient, so that the matrix representation is prefered. However, the dynamical relations of the displacements and the forces at every connecting points are not easily understood with this method alone. This paper develops the mixed use of matrix and signal-flow diagram which makes possible the straightforward representation and computation of dynamical behaviour of complicated systems. Through this method the force-reflecting electro-hydraulic servomechanism is treated, and some comments from the standpoint of the human engineering are refered to the figure of merits of the bilateral servomechanism design.

View full abstract

A Servo valve presented in this paper has been used for an numerically controlled milling machine in which the servo valve is oprated by pulse width modulation. This servo valve is characterized by a pilot valve which has a very nearly ideal lap of valve port, and high corner frequency of approximately 150 cps at 35 kg/cm² supply pressure in the flow versus current response

View full abstract

Although a lot of theoretical works for the synthesis of control systems for random. input have been performed, there are still several points to be solved for the practical application. The author made attempt to apply the statistical technique to process control systems. In process control systems, the correcting action is delayed because of phase lag of the process, so that it is permitted in synthesizing to assume a process control system as a second order system combined with dead time element. In advance of the determination of the compensating network, therefore, designers can estimate the above assumed system through the rms value of the controlled variable calculated by the graph especially prepared for this treatment and then determine the closed loop transfer function suitable to their purpose. The control system, therefore, can be synthesized by the compensating network derived from the closed loop transfer function mentioned above.

View full abstract