Hello,

I am a graduate student in EE currently going through a power electronics class. My focus as an undergrad has been chip design, so not much in the way of power until now.

My question is in the use of SMPS (like buck, boost, etc) to correct power factor.
If I were to describe power factor correction in a nut shell, it is to ensure both the voltage and current wave forms are as close to a sinusoid as possible, and in phase as much as possible. I understand the issues in power factor brought about by bridge rectifiers, and the regulations put in place to limit harmonics in the grid, but obviously the switching adds harmonics right??

This is really my confusion. The SMPS's are added to get a 50/60 Hz waveform to resemble their fundamental sine waves, but a SMPS is switching at, say, 10kHz. If I did an FFT on the output current, wouldn't I see a spike at the switching frequency, plus all of its harmonics? If so, does nobody care about it?

It seems all of the searching on this question always comes back to the PFC in regards to the grid frequency. However, I am more interested in the harmonics injected into DC systems from DC/DC converters. Aren't these high frequencies harmful to other components on the bus as well?

Thanks for the help!

 

PFC refers to the input waveforms. The simplest PFC I have come across is a large - approx 2" cube - iron cored inductor in series with the mains input.

The ripple/spikes on the DC output is separate issue. Because it is high frequency both the inductors and capacitors used to smooth the output voltage can be fairly small. The PCB layout needs to very carefully done to avoid switching currents flowing in any tracks other than those within the SMPS itself, and to avoid radiation which may be picked up by other parts of the circuit. Even quite a short track can develop considerable voltage across it because the inductance of track becomes significant. However well designed the layout and smoothing is there will be some ripple and spikes at the switching frequency (and harmonics of course). For many circuits this won't matter but if the circuit uses low signal voltages and is designed to be sensitive to higher frequencies, e.g. a radio receiver, then techniques to minimise the effects of such possible interference are used. Selecting the frequency of switching, extra local supply filtering, or perhaps using a switcher which uses a spread spectrum technique.
http://www.digikey.com/en/articles/...chniques-to-manage-switching-power-supply-emi

 

Thanks for the reply, this definitely helps.

Although, I am concerned with spikes on the DC side as well as the source side (DC or AC).

For instance, in the design of an AC/DC converter where a buck converter is used for PFC reasons. The switch of the buck is connected to the source side. Aren't the spikes at the switching frequency (and harmonics) injected into the source, in this case the grid?

 

Aren't the spikes at the switching frequency (and harmonics) injected into the source, in this case the grid?

All such supplies have filtering on the input side. This prevents interfering signals from the grid getting in and upsetting the PSU, but is also required by regulation to prevent interference getting out. EMC regulations require that equipment shall not interfere with other equipment and shall be protected against interference from outside.

 

Fantastic, thank you!