Switching Loss Calculator

Turn-on/turn-off crossover energy, switching power at frequency, and total loss including reverse recovery.
Switching Loss
Loss vs Frequency

Crossover Energy & Power

Esw = ½ × V × I × (tr+tf)  •  Psw = Esw × fsw
400V,10A,30/40ns,100kHz
48V,20A,15/20ns,300kHz
Same + Qrr=2µC diode
V
A
ns
ns
Hz
µC
Enter values and press Calculate.

Switching Loss vs Frequency

Uses the same V, I, tr, tf as the calculator above — switch tabs, run a calculation, then view how loss scales with frequency.

Psw vs Switching Frequency

How Switching Loss Happens

Every time a power switch turns on or off, it briefly carries both significant voltage and significant current at the same time as they cross over — and V×I during that crossover is instantaneous power, dissipated as heat. Because this happens on every switching cycle, the resulting switching loss grows directly with frequency, unlike conduction loss which depends only on RMS current. This is the fundamental power-electronics trade-off: higher switching frequency shrinks passive components (see the Magnetics section) but raises switching loss.

QuantityFormula
Crossover energy per eventEsw = ½×V×I×(tr+tf)
Switching powerPsw = Esw × fsw
Reverse recovery lossPrr = Qrr × V × fsw
Total switching-related lossPsw(total) = Psw + Prr

This model uses the simplified linear crossover approximation — treating voltage and current as ramping linearly during tr and tf — which is the standard first-pass estimate used across the industry. Real waveforms are affected by parasitic inductance, gate drive strength, and the device's own capacitances, so measured loss is often somewhat different; this calculator gives the right order of magnitude and correct frequency scaling.

Real-World Applications & Examples

Worked examples

1. 400 V, 10 A MOSFET, 100 kHz. tr=30 ns, tf=40 ns: Esw=½×400×10×70e-9=140 µJ. Psw=140µJ×100000=14 W.
2. Faster gate drive. Halving tr+tf to 35 ns halves Esw to 70 µJ and Psw to 7 W — the direct benefit of a stronger gate driver.
3. 48 V, 20 A, 300 kHz. tr=15 ns, tf=20 ns: Esw=½×48×20×35e-9=16.8 µJ; Psw=16.8µJ×300000=5.04 W.
4. Frequency doubling. Doubling fsw from 100 kHz to 200 kHz in example 1 doubles Psw to 28 W for the same device — switching loss scales linearly with frequency.
5. Adding reverse recovery. A 2 µC Qrr diode at 400 V, 100 kHz adds Prr=2e-6×400×100000=80 W — often larger than the MOSFET's own switching loss, showing why diode selection matters.
6. SiC vs silicon. A SiC MOSFET with 10 ns crossover instead of 70 ns at the same V, I, f cuts switching loss to about 1/7th — the core reason SiC enables much higher switching frequencies.

Frequently Asked Questions

What is switching loss?

It is the energy lost every time a power switch turns on or off, caused by voltage and current briefly overlapping during the transition. Unlike conduction loss, it occurs once per switching cycle and so scales directly with switching frequency.

What is the switching energy (Esw) formula?

The linear-crossover approximation gives Esw = ½×V×I×(tr+tf), where V and I are the switched voltage and current, and tr, tf are the rise and fall (crossover) times.

How do I convert switching energy to power?

Multiply the per-event energy by the switching frequency: Psw = Esw×fsw, since the switch produces one such energy loss per cycle.

Why does higher frequency increase switching loss?

Each switching transition wastes a fixed amount of energy (for given V, I, and speed), so switching more often per second directly multiplies the total power lost — this is the fundamental trade-off against smaller magnetics at high frequency.

How do I reduce switching loss?

Use a stronger/faster gate driver to shorten tr and tf, choose a device with lower capacitances (or a wide-bandgap device like SiC/GaN), reduce the switched voltage or current where possible, or use soft-switching (ZVS/ZCS) topologies that largely eliminate the overlap.

What is reverse recovery loss?

When a diode turns off, stored charge (Qrr) must be swept out before it blocks voltage, causing a brief reverse current spike. This dissipates Prr=Qrr×V×fsw and can rival or exceed the switch's own switching loss.

Why do SiC and GaN devices have much lower switching loss?

They have far smaller parasitic capacitances and, for SiC diodes, negligible reverse recovery charge, so their crossover times are a fraction of silicon's — letting designers push to much higher frequencies for the same loss budget.

How accurate is the linear crossover model?

It captures the correct order of magnitude and the linear frequency scaling that matters for design trade-offs. Real switching is nonlinear (affected by Miller plateau, parasitic inductance and gate drive), so datasheet or measured Eon/Eoff values are more accurate when available.

What is the difference between hard switching and soft switching?

Hard switching (what this calculator models) has full voltage and current overlapping at each transition. Soft-switching topologies (like the LLC resonant converter) arrange for voltage or current to be near zero during the transition, dramatically cutting switching loss.

Should I use datasheet Eon/Eoff instead of this formula?

If the manufacturer provides tested Eon and Eoff values at your operating conditions, those are more accurate than the linear approximation and should be preferred; use this calculator for early estimates or when only rise/fall times are known.

Does gate resistance affect switching loss?

Yes, directly. A larger gate resistor slows the gate charge/discharge, increasing tr and tf and therefore switching loss — see the MOSFET Gate Resistor calculator for that trade-off against ringing and EMI.

How does switching loss combine with conduction loss?

They simply add: total device loss = conduction loss + switching loss (+ any reverse recovery loss from an associated diode). See the MOSFET Power Loss calculator for a combined breakdown specific to MOSFETs.

Related Calculators

MOSFET Power LossMOSFET Gate ResistorJunction TemperatureAll Calculators