What is the difference between CCM, DCM and CRM boost PFC?

In CCM the inductor current never reaches zero, giving low ripple and suiting higher power. In DCM the current falls to zero for part of each cycle, giving high peak currents, so it is used only at low power. In CRM or boundary mode each cycle begins exactly when the current hits zero, allowing soft switching, popular for low-to-medium power at a variable switching frequency.

Why does the duty cycle change over the line cycle?

Because the output bus is fixed while the rectified input swings from zero up to its peak. From D = 1 minus the ratio of input to output voltage, the duty is near 1 when the input is near zero, the largest step-up, and at its smallest when the input is at its peak.

Boost PFC Converter: Working, Waveforms & Formula

Q: What is a boost PFC converter?

It is a boost converter placed after a diode bridge and controlled so that the current in its inductor follows the shape of the rectified line voltage. That shapes the input current into a sine wave in phase with the voltage for near-unity power factor and low harmonics, while raising the output to a fixed DC bus.

Q: Why is a boost converter used for PFC instead of a buck?

Because the boost inductor is at the input, its current is continuous and can be shaped through the whole line cycle including near the zero crossings. A buck cannot draw current when the input voltage is below its output, so it would leave gaps in the input current near every zero crossing and could not shape a clean sine.

Q: Why is the boost PFC output around 400 V?

A boost can only step voltage up, so its output must stay above the peak of the input at all times. For a universal-input supply up to 265 V AC, about 373 V peak, the bus is set around 390 to 400 V to keep a safe margin above that peak.

Q: Does a boost PFC provide isolation?

No. A boost PFC only corrects the power factor and pre-regulates the DC bus; it is not isolated. A safe low-voltage output still needs an isolated DC-DC stage such as an LLC or flyback converter after the PFC.

Introduction — the Power Factor Problem

Most electronic gear runs on DC, so it starts by rectifying the mains with a diode bridge and smoothing it with a big capacitor. It works, but it has an ugly side effect: the capacitor only recharges during the short moment when the line voltage climbs above the voltage already stored on it — right at the top of each half-cycle. So the supply draws current in tall, narrow spikes instead of a smooth sine.

Those spikes are a problem. They make the power factor poor (often only 0.5–0.65), they pump harmonic currents back into the wiring, and they overload the neutral in three-phase installations. Standards such as IEC 61000-3-2 limit these harmonics, so almost every power supply above about 75 W now has to actively clean up its input current. The most popular way to do that is the boost PFC.

What is Power Factor Correction?

Power factor (PF) measures how well the current drawn by a load matches the voltage feeding it — both in timing (are they in phase?) and in shape (is the current a clean sine?). A pure resistor draws a current that is a perfect in-phase sine, so it has PF = 1. The spiky current from a diode-bridge supply is neither in phase nor sine-shaped, so its PF is low and it carries a lot of harmonics (its total harmonic distortion, or THD, can be over 100 %).

Active power factor correction uses a switching converter to force the input current to become a sine wave, lined up with the voltage. Done well, it pushes the power factor above 0.99 and cuts THD below 5 %, so the supply looks like a clean resistive load to the grid.

What is a Boost PFC Converter?

A boost PFC puts a boost (step-up) converter right after the diode bridge and controls it in a special way. Instead of only regulating its output, the controller also forces the current in the boost inductor to follow the shape of the rectified line voltage. Because that inductor sits in series with the input, shaping its current shapes the current drawn from the mains — turning the spiky waveform into a smooth sine that is in phase with the voltage.

At the same time the boost stage lifts the rectified input up to a fixed DC bus, typically around 390–400 V, which the next stage (usually an isolated DC-DC converter) then converts to the final output. So a boost PFC does two jobs at once: it corrects the power factor and it pre-regulates the DC bus.

Boost PFC block diagram: the AC line passes through an EMI filter and a diode bridge into a boost PFC stage of inductor, switch, diode and capacitor, which feeds a roughly 400 volt DC bus and load. A PFC controller senses the input voltage, inductor current and output voltage and drives the switch to shape the input current into a sine.

Figure 1: Boost PFC front-end — the AC line is filtered, rectified, then shaped and boosted to a ~400 V DC bus, with the PFC controller forcing the input current into a sine.

Why the Boost Topology?

The boost converter is the natural choice for PFC for a few practical reasons:

Continuous input current. The inductor is right at the input, so current flows into the converter throughout the whole line cycle. That smooth, controllable current is exactly what you need to shape into a sine.
It works near the zero crossing. Twice per cycle the line voltage passes through zero. A boost can keep drawing and shaping current there because it only ever steps up; a buck would simply stop conducting whenever the input fell below its output.
Simple power path. Just one inductor, one switch, one diode and the output capacitor — cheap and efficient.

The one catch is built into the name: a boost can only step up, so the DC-bus voltage must always sit above the peak of the line (that is why a universal-input supply uses roughly 400 V, comfortably above the ~373 V peak of a 265 V line).

Circuit Diagram & Construction

A boost PFC is built from a rectifier plus a boost stage and its controller:

Diode bridge — rectifies the AC line into a pulsating (rectified-sine) voltage.
Boost inductor (L) — the current-shaping element, carrying the input current.
Switch (MOSFET, Q) — switched on and off fast (tens to hundreds of kHz) to control the inductor current.
Boost diode (D) — passes the inductor's energy to the output when the switch is off.
Output capacitor (C) — holds up the DC bus and supplies the load between pulses.
PFC controller — senses the rectified voltage (as a shape reference), the inductor current, and the output voltage, and sets the switch duty every cycle.

Boost PFC power circuit: diode bridge feeding a boost inductor, MOSFET switch to ground, boost diode to the output capacitor and load, with inductor-current and output-voltage sensing into a PFC controller driving the switch

Figure 2: Boost PFC power circuit — bridge, inductor L, switch Q, diode D, capacitor C and controller

How It Works (Switch On / Off)

The boost stage switches many times within every line half-cycle. Each switching cycle has two states:

Switch ON: the inductor is connected across the rectified input. Its current ramps up, storing energy in the magnetic field. The boost diode is reverse-biased, so the output capacitor alone supplies the load.
Switch OFF: the inductor current must keep flowing, so it forces the boost diode on and pours energy into the output capacitor and load. The current ramps down while the output sits above the input — the “boost”.

The clever part is the control. The controller builds a current reference shaped like the rectified line voltage and scaled by how much power the load needs (from the output-voltage error). It then adjusts the duty cycle every switching cycle so the average inductor current follows that reference. The result over a full line cycle: the input current becomes a sine in phase with the voltage. Notice the duty must be large near the zero crossings (input tiny, big step-up needed) and small near the peak.

The PFC Effect (Input Current Waveforms)

This first set of waveforms is the whole reason boost PFC exists. Compare the current a plain diode-bridge supply draws with the current the same supply draws once a boost PFC is added.

Boost PFC effect: line voltage sine, the spiky non-sinusoidal input current of a diode-bridge supply without PFC, and the smooth sinusoidal input current in phase with the voltage once boost PFC is added.

Figure 3: Without PFC the input current is a pair of tall spikes near each voltage peak; with boost PFC it is a smooth sine, in phase with the line.

Without PFC: current flows only in the brief window when the line exceeds the capacitor voltage, so it appears as sharp spikes near each peak. The current is nothing like the voltage — low power factor, high harmonics.
With boost PFC: the controller draws current all through the cycle, shaped as a sine and aligned with the voltage. The supply now looks resistive to the grid — PF near 1 and low THD.

Switching-Cycle Waveforms

Zoom in to just a few switching periods (the fast time scale). Here you can see how the switch, the inductor, and the boost diode take turns. This is drawn for continuous conduction (the inductor current never reaches zero).

Figure 4: One PFC switching burst — the inductor charges when the switch is on and discharges through the diode when off. The switch and diode currents are the two halves of the inductor current.

Inductor current i_L: a triangular ripple. It rises while the switch is on (input voltage across L) and falls while the switch is off (L drives the output). In CCM it never touches zero.
Switch current i_Q: equals i_L during the on-time and is zero during the off-time — the switch carries the rising part of the ramp.
Diode current i_D: the mirror image — zero when the switch is on, and equal to i_L (the falling part) when the switch is off. This is the current that actually charges the output.

Line-Cycle Waveforms & Current Shaping

Now zoom back out to a full line cycle (the slow time scale). The fast switching ripple still rides on top, but the average inductor current is deliberately shaped to follow the rectified voltage — that is the whole trick. (The ripple is exaggerated here so you can see it.)

Boost PFC line-cycle waveforms: the rectified input voltage, the inductor current whose average follows a rectified sine with switching ripple on top, the duty cycle which is large near the zero crossings and small at the peaks, and the nearly constant output bus voltage with a small ripple at twice the line frequency.

Figure 5: Over a line cycle the inductor-current average tracks the rectified voltage, the duty cycle is largest at the zero crossings and smallest at the peaks, and the output bus stays nearly constant with a small ripple at twice the line frequency.

Rectified input |v_in|: the diode bridge folds the negative half up, giving two humps per line cycle.
Inductor current i_L: the fast switching ripple rides on a slow rectified-sine average (dashed line). Because the input current is this average, the mains sees a clean sine.
Duty cycle D: follows D = 1 − |v_in|/V_out. It is close to 1 at the zero crossings (input near zero, maximum step-up) and drops to its smallest value at the peaks.
Output V_out: the capacitor holds the bus almost constant. A small ripple remains at twice the line frequency, because single-phase power naturally pulses at 2f.

Conduction Modes (CCM, DCM, CRM)

How the inductor current behaves within each switching cycle defines three modes, each suited to a different power level:

CCM — Continuous Conduction Mode: the inductor current never falls to zero. Ripple and peak currents are low, so CCM suits higher power (roughly above 300 W). It usually uses average-current-mode control at a fixed switching frequency.
DCM — Discontinuous Conduction Mode: the current falls to zero and stays there for part of each cycle. Simple but with high peak currents, so it is used only at low power.
CRM / BCM — Critical / Boundary Mode: the next cycle starts exactly when the current reaches zero. It gives natural valley/zero-current switching and is very popular for low-to-medium power (up to a few hundred watts), at the cost of a variable switching frequency.

Key Parameters & Formulas

Quantity	Formula & Value
Boost voltage relation	V_out = V_in / (1 − D)
Duty cycle over the line cycle	D(t) = 1 − \|v_in(t)\| / V_out
Inductor current ripple	ΔI_L = v_in·D / (L·f_sw)
Power factor	PF = P / (V_rms·I_rms) → ≈ 1 with PFC
Typical DC bus (universal input)	V_out ≈ 390–400 V (above the line peak)
Output ripple frequency	2f (100 Hz on a 50 Hz line)
Typical result	PF > 0.99, THD < 5 %

The duty relation says it all: because V_out is fixed and |v_in| swings from 0 to its peak, the duty D swings from near 1 (at the zero crossing) down to its lowest value (at the peak) — automatically, every line cycle.

Advantages & Disadvantages

Advantages

Near-unity power factor and low THD, so it meets harmonic standards like IEC 61000-3-2.
Continuous, controllable input current that is easy to shape over the whole line cycle.
Regulated DC bus ready for the next stage, and it works over a wide (universal) input range.
Simple, efficient power path — one inductor, switch and diode.

Disadvantages

Step-up only: the bus must sit above the line peak, so the output is a high voltage (~400 V).
An extra power stage adds cost and a little loss (typically 1–2 % efficiency).
No isolation — a following isolated DC-DC stage is still needed for a safe low-voltage output.
Switching EMI and start-up inrush that need filtering and protection.

Applications

Switch-mode power supplies for computers, servers and telecom rectifiers.
LED drivers and lighting ballasts that must meet harmonic limits.
EV on-board chargers and adapters above ~75 W.
Appliances and industrial supplies that draw significant power from a single-phase line.

Frequently Asked Questions – FAQs

It is a boost converter placed after a diode bridge and controlled so that the current in its inductor follows the shape of the rectified line voltage. That shapes the input current into a sine wave in phase with the voltage (near-unity power factor, low harmonics) while also raising the output to a fixed DC bus.

Because the boost inductor is at the input, its current is continuous and can be shaped through the whole line cycle, including near the zero crossings. A buck cannot draw current when the input voltage is below its output, so it would leave gaps in the input current near every zero crossing and could not shape a clean sine.

A boost can only step voltage up, so its output must stay above the peak of the input at all times. For a universal-input supply (up to 265 V AC, ~373 V peak) the bus is set around 390–400 V to keep a safe margin above that peak.

In CCM the inductor current never reaches zero, giving low ripple and suiting higher power. In DCM the current falls to zero for part of each cycle, giving high peak currents, so it is used only at low power. In CRM (boundary mode) each cycle begins exactly when the current hits zero, which allows soft switching and is popular for low-to-medium power at a variable switching frequency.

No. A boost PFC only corrects the power factor and pre-regulates the DC bus; it is not isolated. A safe low-voltage output still needs an isolated DC-DC stage (such as an LLC or flyback converter) after the PFC.

Because the output bus is fixed while the rectified input swings from zero up to its peak. From D = 1 − |v_in|/V_out, the duty is near 1 when the input is near zero (largest step-up needed) and at its smallest when the input is at its peak.

Boost PFC Converter