For a 5 kW real power load targeting a 0.95 power factor (PF) on a 230V single-phase AC line, your PFC converter must be rated for 5.26 kVA apparent power and handle 22.9 Amps RMS continuous input current. The core conversion formula is S (kVA) = P (kW) / PF, yielding 5 / 0.95 = 5.26 kVA, and I = 5263 VA / 230V = 22.88 A. If you are correcting an existing uncorrected diode-bridge rectifier load (typical PF 0.65), the pre-correction apparent power was 7.69 kVA; the active PFC converter reduces the line current from 33.4 A down to 22.9 A, resulting in an exact 31% reduction in I²R distribution losses.
The Core Assumptions Fixing Your PFC Converter Size
What locks in these physical dimensions and electrical ratings? Three rigid assumptions: the nominal RMS voltage, the target displacement power factor, and the phase count. A PFC converter doesn't just 'fix' power; it actively shapes the input current waveform to mirror the input voltage sine wave, minimizing reactive power. According to the US Department of Energy, correcting low power factor reduces utility penalties and frees up transformer capacity, but the physical size of your boost inductor core and the thermal mass of your MOSFET heatsink are dictated entirely by the peak inductor current.
Peak current is calculated as √2 × I_RMS divided by the minimum expected line voltage. If your 230V line sags to 200V during brownouts, your RMS current spikes to 26.3 A to maintain 5 kW output. This 15% current increase fundamentally changes your component selection, forcing you to upsize your inductor wire gauge to avoid core saturation and copper melting.
Voltage and Phase Shift Matrix (±20% Load Range)
Presenting a single-voltage answer as universal is a critical design error. The current demand shifts drastically depending on whether you are plugging into a North American 120V split-phase, a European 230V single-phase, or a 400V three-phase industrial feed. Below is the conversion matrix for a 0.95 target PF across a ±20% load range (4 kW to 6 kW).
| Real Power (kW) | Apparent Power (kVA @ 0.95 PF) | 120V 1-Phase (Amps RMS) | 230V 1-Phase (Amps RMS) | 400V 3-Phase (Amps RMS) |
|---|---|---|---|---|
| 4.0 kW (-20%) | 4.21 kVA | 35.1 A | 18.3 A | 6.1 A |
| 5.0 kW (Baseline) | 5.26 kVA | 43.9 A | 22.9 A | 7.6 A |
| 6.0 kW (+20%) | 6.32 kVA | 52.6 A | 27.5 A | 9.1 A |
Note: 3-Phase current is calculated using I = P / (√3 × V_LL × PF). The massive drop in current on 3-phase systems is why industrial motor drives and EV chargers above 7 kW mandate three-phase inputs.
When PFC Conversion is Meaningless (The Unknown PF Trap)
This math becomes entirely meaningless when the load's baseline Total Harmonic Distortion (THD) or displacement power factor is completely unknown, or when the load is highly dynamic. You cannot size a passive capacitor bank or an active boost stage without knowing the baseline reactive power (Q).
Furthermore, if you are sizing an active PFC for a non-linear load with extreme crest factors (like a cheap switching supply without its own internal PFC, or a pulsed laser driver), the RMS-to-peak ratio breaks the standard sine-wave assumptions. In these scenarios, the basic S = P / PF conversion will yield an RMS current that is dangerously undersized for peak current handling. The Texas Instruments PFC Design Center explicitly warns that high-crest-factor loads require oversizing the PFC stage's current sense resistor and boost diode to handle instantaneous peak currents that can exceed 3x the calculated RMS value.
Decision Tree: Picking the Exact PFC Topology and Part
Stop guessing topologies. Use this decision path to terminate your design phase and select concrete silicon. Match your power range and phase count to lock in the controller and power switches.
| Power Range & Phase | Topology Choice | Controller IC | Power Switch (Concrete Pick) |
|---|---|---|---|
| < 300W, 1-Phase | Boundary Conduction Mode (BCM) | STMicroelectronics L6562A | Infineon IPA60R099P7 (600V Superjunction) |
| 300W - 3kW, 1-Phase | Continuous Conduction Mode (CCM) | Texas Instruments UCC28180 | Infineon IMW65R045M1 (650V SiC MOSFET) |
| > 3kW, 1-Phase | Interleaved CCM Boost | TI UCC28070 | Wolfspeed C3M0045120K (1200V SiC) |
| > 3kW, 3-Phase | Vienna Rectifier / 3-ph Boost | TI UCC28070 (Interleaved) | Infineon IMZ120R045M1 (1200V SiC) |
The Final Verdict for our 5 kW 230V Benchmark: You fall into the high-power 1-phase edge case. A single-phase CCM boost at 22.9 A will suffer from massive ripple current and unmanageable inductor sizes. You must use an interleaved topology. Default Pick: Use the TI UCC28070 interleaved controller paired with two Infineon IMW65R045M1 650V SiC MOSFETs to split the thermal load and halve the input ripple current.
FAQ: PFC Converter Sizing Edge Cases
Q: Does a PFC converter change the real power (kW) consumed by the load?
A: No. A PFC converter only reduces apparent power (kVA) and reactive current. The real power (kW) consumed by the downstream load remains identical; you are simply drawing it more efficiently from the grid, which reduces utility penalty fees and prevents upstream breaker trips.
Q: What happens if I size the PFC converter for 0.99 PF instead of 0.95?
A: You hit the law of diminishing returns. Pushing from 0.95 to 0.99 PF requires significantly tighter current loop bandwidth, a larger boost inductor to maintain continuous conduction at zero-crossings, and a much more expensive controller. For 95% of industrial and commercial applications, 0.95 PF is the sweet spot that avoids utility penalties without over-engineering the hardware.
Q: Can I use a standard silicon MOSFET for a 5 kW PFC stage?
A: Technically yes, but practically no. At 5 kW, the switching losses in a standard silicon Superjunction MOSFET operating at 100kHz in CCM mode will require a massive heatsink. Switching to 650V Silicon Carbide (SiC) MOSFETs eliminates reverse recovery losses in the boost diode and cuts switching losses by up to 60%, allowing you to shrink the magnetic components and thermal management hardware.
