Hi. Can someone explain to me how a quasi square current results in a dc link? I can't get my head round it.

 

The link current is usually constrained by a series inductance to be an approximately constant (continuous) DC value at any given load condition. There will be some ripple on the DC depending on the power level. If the rectifier system feeding the link is phase controlled, the ripple would increase with increasing firing delay angle [α].

With a constant DC link current, the rectifier currents are then notionally square pulses of duration determined by the rectifier pulse number. Normally one would expect to see a poly-phase rectifier configuration. With a three phase full-wave rectifier for instance, the individual rectifier element current pulses are of 120° (electrical) duration with amplitude equal to the steady state DC link current.

The rectifier AC side line currents are then not pure sinusoidal waveforms but rather quasi (or modified) square wave currents approaching an approximate sinusoidal form. The squareness of these current waveforms means they have a relatively high harmonic content - which is undesirable but unavoidable.

The only way to fully appreciate this is to draw a time based waveform diagram of the various currents in the system - noting the effect of superimposing the various rectifier element currents - which leads to the "construction" of the effective AC line currents. For the purpose of such an exercise one would start with the assumption of a perfectly constant DC link current.