Active PFC Power Supplies: What They Are and Why Your UPS Cares

The problem PFC solves

Every switch-mode power supply starts by rectifying AC mains into DC and storing it on a large capacitor. The simplest way to do that is a bridge rectifier feeding the capacitor directly. It works, but it has an ugly side effect: the capacitor only accepts current when the incoming voltage rises above the voltage already stored on it, which happens near the peak of each half-cycle.

The result is a current waveform made of narrow, tall pulses instead of a smooth sine wave. Real power gets delivered, but the RMS current is much higher than it needs to be, and the pulses are rich in harmonics (multiples of 60 Hz). The ratio of real power to apparent power, the power factor, comes out around 0.5 to 0.7 for such a supply.

High harmonic current heats building neutrals, distorts the voltage for everyone else on the circuit, and wastes capacity in wiring, transformers and UPS inverters. European limits on harmonic emissions (IEC 61000-3-2, applying to most equipment above 75 W input) pushed manufacturers to fix it, and the fix spread worldwide. For the broader picture of harmonics on a building's power, see harmonics and THD.

Three ways a power supply can draw current

Inside an active PFC stage

The active PFC circuit sits between the rectifier and the bulk capacitor. A controller switches a transistor tens of thousands of times per second, building energy in an inductor and releasing it to the capacitor at a voltage higher than the mains peak, commonly around 380 to 400 V DC. By adjusting the switching duty cycle continuously, the controller makes the average input current track the shape of the input voltage. To the wall, the supply looks almost like a resistor.

A useful consequence: the PFC stage is a feedback loop that expects a sinusoidal input. It senses input voltage, predicts how much current it should draw at each point in the cycle, and regulates hard. That is great on clean utility power. It also explains why an unusual input waveform can confuse it.

Active PFC and the 80 PLUS badge

The 80 PLUS certification program for computer power supplies tests efficiency at several load points, and its requirements include a power factor of at least 0.9 at 100% of rated load. In practice only active PFC designs reach that reliably, so an 80 PLUS badge (any level, from the base tier up to Titanium) is strong evidence of active PFC.

The efficiency tiers themselves (Bronze, Gold, Platinum and so on) describe how much of the input power reaches the components as DC rather than becoming heat. They are about efficiency, not power factor beyond that 0.9 threshold. A Gold supply is not "more PFC" than a Bronze one.

Why simulated sine wave output can trip active PFC supplies

Standby and many line-interactive UPS units produce a simulated (stepped) sine wave on battery: the voltage jumps to a positive level, holds flat, drops to zero for a pause, then jumps negative. Its RMS value matches mains, but its shape is very different. Reported symptoms with some active PFC supplies include:

• The computer shuts off or reboots the instant the UPS transfers to battery.
• The computer survives transfer but turns off seconds later, typically under gaming or rendering load.
• The UPS itself reports an overload or shuts down although the measured wattage is well within rating.
• An audible buzz from the PSU while on battery.

The exact mechanism differs between PSU designs and is not fully documented by manufacturers. Several contributing factors are widely discussed by power electronics engineers:

1. Peak voltage and the flat top. Depending on the UPS design, the stepped wave's peak may be lower than the roughly 170 V peak of a true 120 V sine wave, and its pulse width often changes as the battery discharges. The flat top gives the PFC controller an input shape it was not designed to track, and a lower peak means it must work harder to keep its bulk capacitor charged.
2. The steep edges. At each step the voltage changes very quickly. Input filter capacitors and the PFC inductor respond with current spikes that the UPS inverter sees as surges. If those spikes hit the inverter's current limit, the UPS may declare overload.
3. The zero-voltage dwell. During the pause at zero, the PFC stage has nothing to draw from. Some controllers interpret a long enough gap as brownout or line loss and shut down to protect themselves.
4. Transfer gap plus hold-up. The PSU must also ride through the UPS's transfer time on its bulk capacitor. Desktop supplies generally hold up for something on the order of 10 to 20 ms at full load, less with a heavily loaded supply, so a slow transfer combined with a confused restart of the PFC loop is a plausible failure path at high load.

None of these guarantees a failure. Plenty of active PFC computers run fine on stepped output at light to moderate load. The risk rises with load percentage on both the PSU and the UPS, which is why "it worked when I tested it at the desktop" is not proof it will survive a gaming session in an outage.

How to tell whether your PSU has active PFC

1. Read the label. Most PSUs print "Active PFC" or a power factor value on the side label or in the specifications. Server supplies list power factor at several loads.
2. Look for an 80 PLUS badge. Any level implies a power factor of 0.9 or better at full load, which in practice means active PFC.
3. Check the input range. "100-240 V~" with no red voltage selector switch on the back is typical of active PFC. A 115/230 V switch usually means passive or no PFC.
4. Measure it. Plug-in energy meters often display power factor. A reading of 0.95 or higher under moderate load means active PFC; readings in the 0.5 to 0.8 range mean passive or none. Our page on measuring power draw explains how to read these meters.
5. Look up the model. Manufacturer spec sheets and reputable PSU reviews list PFC type and measured power factor.

Worked example: does PFC change the UPS you need?

Two otherwise identical office PCs each draw 200 W of real power. One has an older passive PFC supply measured at PF 0.70; the other has an active PFC supply at PF 0.98.

Put four of the passive PCs on a 1500 VA / 900 W UPS: 800 W is 89% of the watt rating and 1,144 VA is 76% of the VA rating. Both are high, and the watt limit is the binding one. Four active PFC PCs: still 800 W (89%) and about 816 VA (54%). PFC lowered the VA load sharply but did nothing for watts, and watts is what drains the battery. The lesson matches the VA vs watts guide: with modern loads, compare UPS units by watts.

The waveform question is separate. The active PFC fleet should go on a pure sine wave UPS; the passive fleet would tolerate either waveform. For gaming builds, the gaming PC sizing guide adds GPU transient headroom on top of this.

Testing an existing setup safely

If you already own a simulated sine wave UPS and an active PFC computer, a controlled test tells you more than any compatibility list:

1. Save your work and close anything important. Make sure the UPS battery is fully charged.
2. Put the PC under its typical heavy load (the game or workload you actually run).
3. Unplug the UPS input cord from the wall (or switch off its breaker if that is easier). Do not just flip the UPS power switch; that turns off the output.
4. Watch for at least 60 seconds. Note any reboot, shutdown, buzz or UPS alarm.
5. Plug the UPS back in and confirm a clean return to utility power.

If the PC shuts off, the fixes are a pure sine wave UPS, a larger UPS to reduce load percentage (partial and unreliable), or a different PSU. The troubleshooting page UPS shuts off during an outage covers the other causes, such as a weak battery, that produce the same symptom.