AKKN ElectronicsMetallized polypropylene snubber capacitors absorb the energy stored in the stray inductance of an IGBT half-bridge at every hard switching transition. That energy, 0.5 × Lstray × I², is discharged into the capacitor thousands of times per second, producing a repetitive ripple current that heats the dielectric. Combined with the DC bus voltage, this determines the useful service life of the part. Snubber selection is therefore a matter of managing heat and voltage derating, not simply picking a capacitance value.
How Snubber Capacitor Life Is Usually Modeled
Film capacitors do not contain an electrolyte, so their aging mechanisms differ from aluminum electrolytic units. Wear-out comes from dielectric degradation and occasional self-healing clears. Two stress factors dominate the life calculation.
- Hot-spot temperature. Life roughly halves for every 10 °C increase above a reference temperature, and doubles for every 10 °C decrease. Most metallized polypropylene series use a reference hot-spot temperature of 55 °C and carry a category temperature limit of 85–105 °C.
- Applied DC voltage. Life decreases according to a power-law of the applied voltage. Representative exponents are between 4 and 7. Operating at 75 % of rated voltage instead of 90 % typically multiplies expected life by about 2 to 4.
The hot-spot temperature is the sum of ambient temperature and self-heating: Thot = Tamb + Rth × (Irms² × ESR). For a boxed radial-lead part in the 0.1–1 µF class, Rth is typically 20–30 K/W, and ESR at 10 kHz is on the order of 5–15 mΩ.
Ripple-Current Derating: What the Numbers Say
A snubber capacitor carries a pulse of current at each switching edge rather than a steady waveform, but the heating effect is governed by the rms value. In a 2–20 kHz IGBT inverter, a 0.47 µF snubber typically sees 5–15 A rms depending on stray inductance and switching speed.
Consider a generic 0.47 µF, 850 V rated part with 10 mΩ ESR and 25 K/W thermal resistance. At 10 A rms, dissipation is 1 W and the hot-spot rises 25 K above ambient. In a 50 °C cabinet, the hot-spot reaches 75 °C, where expected life falls to roughly 25,000 h. Reduce ripple to 6 A: dissipation drops to 0.36 W, the rise to 9 K, and the hot-spot stays near 59 °C, extending life to about 80,000 h.
The table below applies the same model across duty conditions for that part, operating at a 600 V bus (71 % of rated voltage) with a reference life of 100,000 h at 55 °C hot-spot.
| Duty condition | Ambient (°C) | RMS ripple (A) | Self-heating (K) | Hot-spot (°C) | Expected life (h) |
|---|---|---|---|---|---|
| Light | 40 | 5 | 6 | 46 | > 100,000 |
| Moderate | 50 | 8 | 16 | 66 | ~ 50,000 |
| Heavy | 55 | 12 | 36 | 91 | ~ 9,000 |
| Extreme | 60 | 15 | 56 | 116 | Not acceptable — above category limit |
Note that a modest reduction in ripple current buys a disproportionately longer life, because the temperature penalty is quadratic in the current.
Design Rules for Snubber Capacitor Selection
- Keep the DC bus voltage at or below 75 % of rated DC voltage. For a 600 V bus, select an 850 V rated part; for an 830 V bus, use a 1200 V rated part.
- Estimate self-heating before choosing capacitance: P ≈ 0.5 × C × ΔV² × f, where ΔV is the clamping voltage swing across the snubber per transition (typically 60–150 V in a 600 V system) and f is the switching frequency.
- Hold the hot-spot at least 20 K below the category temperature limit of the film series. If a prototype runs hotter, reduce ripple by lowering stray inductance, or select a larger-body part with lower Rth, rather than increasing capacitance.
- Prefer a part with dissipation factor below 0.0005 at 10 kHz. Low DF keeps ESR and therefore heating down.
- Do not oversize capacitance arbitrarily. A larger C reduces ΔV but increases the energy dissipated per cycle at the same switching frequency, which raises the hot-spot temperature. In a typical 50–200 A IGBT module, 0.1–1.0 µF per half-bridge covers most designs.
- Keep the snubber loop inductance below 30 nH. Each additional 10 nH in the loop raises the turn-off overshoot voltage by approximately L × di/dt and forces the snubber to work harder.
In practice, verify the model once in a prototype: measure the surface temperature of the mounted snubber at full load; the surface runs 5–15 K below the internal hot-spot. Use that reading to confirm the derating decision. When an original snubber part is obsolete, matching capacitance, rated voltage, dissipation factor, body dimensions, and the operating ripple current at your switching frequency is generally sufficient to identify a cross-reference with equivalent lifetime.
