In a previous article I discussed the problem of poor power factor in off-line switched-mode power supplies. We showed that the current waveform of a typical circuit where the mains is directly rectified and filtered looks very spikey (Figure 1, top) because the rectifier diodes only conduct to charge the filter capacitor at the peaks of the mains cycle.
A typical SMPS directly rectifies and filters the mains resulting in a spikey current waveform as shown at the top. This has poor power factor due to the presence of high levels of current harmonics. For close to unity power factor, we need the current waveform to be near-sinusoidal (and in phase with the voltage) with correspondingly low harmonics.
The poor power factor is due to the high harmonic content of the current waveform. Figure 1 also shows the harmonic content of a typical rectifier-filter current waveform. You can see the harmonic content in our example is significant up to the 13th harmonic. For unity power factor (the ideal case shown at the bottom of the figure) , the current waveform should be near-sinusoidal and therefore have a minimum of harmonics beyond the fundamental frequency.
The simplest way to improve power factor is to add an inductor in series with the circuit as shown in Figure 2. If it is large enough, the inductor has the effect of filtering out some of the higher harmonics, making the current waveform more sinusoidal in shape. There are however a couple of big disadvantages to this passive approach.
A simple way to minimise harmonics in such a circuit is to introduce a large inductor. Unfortunately, this inductor would be impractically large and expensive for most applications, hence the adoption of active power factor correction circuits.
First, the size and weight of an inductor that will make an appreciable improvement more or less obviates any advantage of using a switch-mode power supply in the first place. Secondly, it is hard to design such a circuit to efficiently handle a wide range on input voltages. Typical “universal” power supplies can operate over a 90 to 250V range at 50 or 60Hz so can be used anywhere in the world.
This leaves us with active approaches. The simplest, and most common for small power ratings (say under 200W), is to add a boost converter between the bridge rectifier and the filter capacitors. This is illustrated in Figure 3. The boost converter’s input is the rectified but unfiltered mains. When the MOSFET conducts, the boost diode is reverse biased, and the current drawn from the mains ramps up in the inductor until some threshold is reached. The MOSFET is then turned off and the current in the inductor flows through the diode charging the filter capacitor. The inductor current ramps down (usually to zero) during this phase. This switching repeats at a frequency much higher than the mains frequency (typically 50kHz or more).
The typical boost PFC circuit introduces a boost converter between the rectifier and filter as shown. The converter is modulated such that the average current – shown here dotted in red) follows a near sinusoidal trajectory.
The clever part is that the peak current threshold is modulated to track the shape of the input voltage waveform (a rectified sine wave), so the average mains current follows a sinusoidal shape shown dotted in the figure. Since the switching occurs at a high frequency, the boost inductor can be relatively small, and the high frequency current ripple is easy to filter.
For higher power circuits it is common to use two boost converters in interleaved mode (Figure 4). Here the two converters are operated 180˚ out of phase. In this case the switching frequency ripple currents in the inductors cancel each other out and the resulting mains ripple is very low.
For higher power levels a second boost converter is often used in the “interleaved: topology shown here. The two converters operate 180˚ out of phase so that the ripple current in the two inductors cancel out, resulting in very low mains current ripple.
Of course, the boost converter approach does introduce a level of inefficiency, since we have added an additional diode drop in series with the bridge rectifier. If efficiency is your biggest concern, you can use a “bridgeless” circuit like that in Figure 5a. Here the functions of rectifier and boost diode are integrated. You don’t have to use two inductors as shown here, but it does help reduce the noise pushed back onto the mains by the high dV/dt switching of the MOSFETs.
To improve efficiency, one could adopt one of the two PFC topologies shown here. These combine the rectifier and boost diodes to eliminate one set of voltage drops but come with some additional complexity in the control circuit.
Another popular choice is the “totem pole” topology shown in Figure 5b. In this case one MOSFET acts as a synchronous rectifier each half-cycle, while the other is switching. This can reduce losses and even further than the bridgeless approach since the voltage drop across the MOSFET is much lower than that across a diode.
Simple one-chip solutions are available from many manufacturers to make designing one of these circuits easier. Figure 6 shows one example of a practical circuit taken from a TI app note. This is a 100W booster using the UC3853. One clever trick used here is to power the chip via an auxiliary winding on the boost inductor.
This example form TI shows a practical 100W boost PFC circuit to give you an idea of what’s involved. Note the clever use of an auxiliary winding on the boost inductor to power the control chip.
You should definitely consider using some kind of power factor correction circuit on any switched-mode power supplies you might design, especially if the power is above a few hundred watts.
Bibliography
“Power Factor Correction (PFC) Circuits,” 2019. https://toshiba.semicon-storage.com/info/application_note_en_20191106_AKX00080.pdf?did=68570.
Todd, Phillip C. “Boost_Power_Factor_Corrector_Design_with_the_UC3853,” n.d. https://www.ti.com/lit/an/slua080/slua080.pdf.
Sponsor this Article
Andrew Levido (andrew.levido@gmail.com) earned a bachelor’s degree in Electrical Engineering in Sydney, Australia, in 1986. He worked for several years in R&D for power electronics and telecommunication companies before moving into management roles. Andrew has maintained a hands-on interest in electronics, particularly embedded systems, power electronics, and control theory in his free time. Over the years he has written a number of articles for various electronics publications and occasionally provides consulting services as time allows.
