Search

A method to determine the small-signal control-to-output transfer function of quasi-resonant converters operating with continuous current is presented. The design of the feedback loop and compensation network becomes easier by this approach. Investigation is carried out using the boost converter topology.

An algorithm for open-loop model matching is introduced. The designed open-loop transfer function is shown to match a prespecified set of Pade coefficients and Markov parameters. The algorithm can be used in the design of robust controllers.

An active C filter realizing the general biquadratic transfer function is given, using signal flow graphs.

The method of moments is used to fit a pulse transfer function of given order r to a given sampled impulse-response function. By this means, the model correctly predicts the system response to an arbitrary polynomial input of order 2r?1, and correctly predicts the required polynomial input to correct for additive polynomial drift at the output of order 2r?1.

The letter describes a circuit designed to operate from a ±12 V power supply over a 7:1 range of input levels, on the principle of a diode function generator whose transfer characteristic, when phase-inverted and added to a constant level, approximates to a reciprocal curve. Diode intrinsic non-linearity significantly improves the smoothness of the transfer function.

The problem of the simultaneous realisation of a transfer function and a transfer-function sensitivity using a single feedback path is discussed. The results are illustrated by a simple example.

Loop-bandwidth adaptation with varying noise level can be obtained by including a device with a modified limiter transfer function in the loop of a phase-lock loop. A simple realisation of the transfer function and the change in p.l.l. performance are described. Flexibility advantages in use are obtained by this approach.

A new procedure is given for the synthesis of voltage transfer functions. It is shown that the well known circuits of Lovering and Brugler can also be derived easily by this method. It is also indicated that a simplification can be obtained in the RC:-RC decomposition method. which is in the proposed procedure.