An ‘ideal* converter would accept the power flow of a 3-phase a.c. system operating with sinusoidal voltage and current, and, without energy storage and by a continuous process, convert to or from d.c. Present-day converters rely, however, on repetitive circuit switching operations, more than 12 per cycle being generally uneconomic despite the cost of the energy storage components required in damping circuits and in the filters to maintain acceptable waveforms. Analysis of the operation of such converters is based on the mathematics of repetitive transients (Laplace and Fourier) and on the use of a d.c. transmission simulator, an extensive model at 10
-7
scale in power, which is also necessary in the development of complex electronic control circuits. There exists a great background of experience contributing to the design of most components of the power circuit. In contrast, the development of the switching device, whether thyristor stack or mercury arc valve, calls for advances in the state of art, both in scientific appreciation and in technology, which must be supported by full scale tests. There is little immediate prospect of the theoretical ‘ ideal * converter, but this is unimportant, provided that development leads to enhanced overall reliability.
Fast online diagnosis of open-circuit switching faults in flying capacitor multilevel inverters