# How does the PFC boost circuit work?

I am reading TI's PFC Circuit Basics, and I am having a hard time understanding how the active PFC boost circuit works. I understand that ordinarily the current waveform is not in-phase with the voltage waveform and that PFC corrects that somehow, but I can't fully visualize the waveform of the current flowing through the inductor.

What does the actual waveform look like?

I saw this figure that shows inductor ripple current, so I think maybe this could be what it looks like (almost like a boost converter ripple current that follows the sinusoidal arc of the input voltage.) It is also unclear to me how the voltage gets boosted as well - maybe if I had a good simulation to look at I could understand better.

• As explained here, there are two distinct ways of having a poor power factor, reactive loads (phase-shifted current waveform) and nonlinear loads (non-sinusoidal current waveform). Power supplies fall into the latter category, and this is what the PFC circuit is fixing. Aug 20, 2022 at 21:19
• Is your lack of understanding in how a regular old boost converter works, or how a PFC boost converter works? A PFC boost is basically a regular old boost converter that's controlled such that the current it draws from the line is sinusoidal and in phase with the line voltage. Aug 20, 2022 at 21:41
• Hey Tim, yeah I guess it is not intuitive to me how the 1) Making sure the current is in-phase with the line voltage and the 2) regulating a constant output voltage are both accomplished at the same time. Aug 20, 2022 at 21:45

To visualize the inductor current in a PFC inductor, the best is to build a prototype or run a cycle-by-cycle simulation. The input voltage will be the grid sine wave and a dedicated circuitry will actuate a power switch on and off according to a control law. For a boost PFC working in a self-relaxing (no clock) boundary mode conduction (abbreviated BCM or CrM for critical conduction mode), the controller maintains a constant on-time along the input period to deliver nominal power and it naturally performs power factor correction. The frequency is changing depending where you are on the sine wave because magnetization and demagnetization times vary with $$\V_{in}\$$.

To look at the PFC inductor current, you can resort to an equation-based graph, build a prototype or simply run a cycle-by-cycle simulation. The one below is excerpted from my 60+ SIMPLIS templates that you can freely download from my webpage:

The simulation time takes 6-7 mn on my machine for a 800-ms run. If you zoom on a 100-ms period of time, you obtain the below graph. The input current is nicely sinusoidal, with some crossover distortion at the 0-V input region. Techniques exist to improve this by artificially increasing the on-time in this area:

You can see how the duty ratio and the frequency evolve along a grid period. I purposely increased the inductor to a few mH to reduce the switching frequency so that you can have a better look at the inductor current:

The averaged input current is the filtered version of the averaged inductor current. In a BCM-operated converter, the average inductor current is the peak current divided by two. Therefore, if the inductor peak current envelope follows a sinusoidal shape, the input current will also be sinusoidal. In voltage-mode control BCM, the controller blindly imposes a fixed on-time (adjusted with $$\P_{out}\$$) without sensing the input voltage (this is the simulated schematic I shown) while in current-mode control, the inductor peak current is set via a scaled-down image of the rectified voltage and the error voltage via a multiplier (see the venerable MC33262 from MOT).

• Okay this makes more sense to me now. It seems like over the course of one cycle (half of a rectified 60Hz/50Hz sine wave) the PFC controller does a lot to accomplish both objectives I laid out in other responses. Thank you! Aug 21, 2022 at 18:29

I understand that ordinarily the current waveform is not in-phase with the voltage waveform and that PFC corrects that

That is for a completely different scenario; that scenario being when you are trying to power factor correct an unknown (usually inductive) load such as a motor.

The PFC circuit in your question is trying to ensure that it creates a high-voltage DC (such as 400 volts) and not do so in a way that creates a lot of harmonic distortion in the current it draws from the AC supply. In short, it tries to present a load to the AC supply that "looks" resistive.

A regular bridge rectifier and smoothing capacitor creates horrendous current distortion and harmonics so, when transfer powers get high, legislation says you must PFC. That is what this is all about.

Then, whatever connects to the PFC's high-voltage DC output can draw whatever load current it wants (within reason) and, it can do so knowing it cannot make the current into the PFC correction circuit appear to be anything else other than resistive.

The two scenarios are different.

I saw this figure that shows inductor ripple current

They are not very accurate at portraying the main subtleties. For instance, the falling slope of the current will always be a more constant rate because, the output voltage of the inductor is supplying (or topping up) the DC supply and, that DC supply (once established) is around 400 volts and constant. Maybe try this (that I drew for another answer): -

And, the main thing to observe is that the green charge lines are varying in slope as the AC voltage rises and falls over each half-cycle of AC. Also observe that the red transfer current slope is more constant (dumbed-down version).

• So, green slope is proportional to the input AC voltage and red slope is less variable in slope.

Of course, as Tim Williams points out in a comment below this answer, it's a bit more complex than this because the red-trace slope will reduce as the input voltage rises towards its peak. The inductor only has to supply the difference between the output DC voltage and the input rectified waveform and, when the input voltage is closer to peaking, the voltage difference between input and DC output is less and thus the slope of the red line is less.

The waveforms above are dumbing-down the truth to make it easier to follow for the uninitiated. Dumbing-down is something that might help in this circumstance for the OP. But, it comes with a price and hopefully, that will be appreciated rather than scorned by people who are "in the know". Maybe this added picture will help (less-dumbed-down): -

It is also unclear to me how the voltage gets boosted as well

If you need to know how boost converters work, then I suggest you ask a question about those before trying to figure out how the front-end PFC circuit works. Of course, they are very-related but, you need to grasp basic boost converter operation first.

• Hi Andy, thanks for your answer. To clarify my question about the voltage getting boosted, I meant that it is not clear to me how 1) regulating the line current (making the load look resistive) and 2) maintaining a constant output voltage are able to be done at the same time. What does the switching behavior of the MOSFET look like (duty cycle)? Aug 20, 2022 at 22:45
• @RGBEngineer you can infer the switching duty cycle from the picture; when the inductor is charging (green) the MOSFET is on; when the inductor is releasing it's stored energy to the output capacitor (red transfer current), the MOSFET is off. The line current (from the AC) is not regulated i.e. it is not constant; the current taken is made proportional to the rectified AC voltage and hence, although it is switching at high frequency, the average shape of the current is a half-sine-wave. Aug 20, 2022 at 22:58
• How are 1) regulating the line current (making the load look resistive) and 2) maintaining a constant output voltage are able to be done at the same time? Aug 20, 2022 at 23:01
• It can be done is the important thing here but, it's tricky to get right (never perfect but good enough). I've previously answered a question you have raised on boost converters and you just have to imagine that the DC input to a boost converter can vary from a low value (a few volts) to a large value (maybe 300 odd volts) and that the controller has to sort-things-out correctly. Of course, it's easy to imagine this if the DC supply changed gradually but, when the DC supply rises from zero to several hundred volts and back to zero in 20 ms, it's a little trickier. Stick at it is my advice. Aug 20, 2022 at 23:06
• @TimWilliams you are absolutely correct and despite me dumbing down the answer to reach the level needed I became unstuck. I shall add a few words to clear this up. Aug 21, 2022 at 0:03

A greatly simplified example of how a PFC converter works is to consider that a wide input range switching supply or DC-DC converter draws more current with the minimum input voltage (typically 95 VAC) and less current with maximum input voltage (typically 265 VAC). If you remove the input storage capacitors, the switching circuit will adjust its PWM over each half-cycle of rectified sine wave, so that it provides the same regulated output voltage, but input current (and PWM) will be highest when the voltage is low, and will be lowest when input voltage is highest. Energy storage, filtering, and regulation will be performed by the switching circuit.

Here is a simulation of a PFC circuit using an LT1249 IC. It shows a comparison between a conventional FWB rectifier circuit and the equivalent using PFC. Note: I just saw that they both used 600 Hz AC input. I'll try again with 50 Hz to see if there is any significant difference.

I had to tweak some circuit components to get a reasonable response, but it looks OK at 50 Hz. Input current surges are definitely lower.