Semiconductor Fuses, I²t and High-Speed Protection

Why semiconductor fuses are different

A semiconductor fuse is selected for the survival of a device that can be destroyed by a very short fault. General-purpose fuse links are designed to protect cables and circuits. A semiconductor fuse is designed to reduce the current peak and the energy passed into a protected device during a severe fault.

This difference matters in power electronics. A rectifier diode, SCR, thyristor, IGBT module or converter bridge may have a withstand limit that is lower than the energy a standard fuse would allow through. The fuse must interrupt the fault before the silicon junction, bond wires, module case, busbar connection or adjacent equipment is damaged.

The normal current rating still matters, but it is only the beginning. The real selection checks are voltage, breaking capacity, time-current curve, pre-arcing I²t, total clearing I²t, peak let-through current, cooling, body style and mounting method.

I²t: the value many replacements miss

I²t is a way of expressing the thermal energy associated with a fault current over time. In semiconductor protection, the fuse data is compared with the published withstand value of the protected device. A fuse can have the same current rating and the same physical size as another fuse, but a different I²t value and a different level of real protection.

Two I²t values are often discussed. Pre-arcing I²t covers the energy before the fuse element melts. Clearing I²t covers the full interruption event, including arcing. For device protection, total clearing behaviour is usually the practical concern because the device sees the fault until the circuit is fully interrupted.

The circuit also changes the result. DC circuits, high available current, cable inductance, capacitor discharge and converter topology can alter the let-through energy. That is why a simple rule such as “same amps, same fuse” is unsafe for power semiconductor applications.

Where semiconductor fuses are used

The most common applications are power conversion circuits. A fuse may be placed on the AC input side of a rectifier, in a DC link, in a branch feeding an inverter, or directly in series with a power semiconductor module. The exact position depends on the equipment design and the failure mode being protected.

Variable-speed drives and soft starters use power semiconductor devices to control motors. UPS systems and battery converters use high-current DC sections where a fault can rise rapidly. PV inverters and EV charger power stages can combine high DC voltage with sensitive switching devices. In all of these circuits, the fuse has to match the real fault path, not only the nameplate load current.

Industrial welders, traction converters, battery energy storage systems and high-power DC supplies are also common applications. In these systems the semiconductor fuse is often chosen because a normal general-purpose fuse may let too much peak current or too much I²t energy through before clearing. Where the protected device has a published withstand limit, the fuse has to be selected so the clearing behaviour stays below that limit under realistic fault conditions.

Semiconductor fuses are also found in rectifier cabinets, DC drives, crane drives, marine converters, railway auxiliary converters, induction heating equipment, large battery chargers and power supplies for process plants. These circuits can contain stored energy in capacitors, strong DC sources or transformer-fed fault currents, so the fuse has to clear fast enough before the semiconductor junction, bonding wires, busbar connections or module case suffer damage.

Placement matters as much as fuse class. A fuse on the AC supply side can help limit fault energy reaching the rectifier or converter, while a fuse in the DC link or close to the module may respond more directly to internal semiconductor faults. Capacitor discharge, low circuit inductance and battery-fed faults can all change the duty, so the safest location is the one confirmed by the equipment design and manufacturer data rather than guesswork.

In a drive or inverter, the same fuse position can protect different parts depending on the fault path. A line-side fuse may protect the rectifier bridge and incoming conductors. A DC-link fuse may limit capacitor or busbar fault energy. A branch fuse close to a converter stack may protect one module while leaving the rest of the equipment easier to inspect and restore. This is why the application should be read as a system, not as a single fuse holder.

Some fuses protect the semiconductor itself. Others protect the branch circuit, cable, busbar or enclosure. A good design often uses more than one protection layer, with the semiconductor fuse coordinated against upstream devices so one fault does not remove more of the system than necessary. In larger systems, this coordination also helps reduce downtime because a faulted converter section can be isolated without taking out the whole feeder or plant section.

For maintenance teams, this is the practical point: the correct semiconductor fuse is usually tied to the equipment design. If a fuse has opened in a UPS, inverter, EV charger, PV inverter or DC drive, the cause should be checked before replacement. Replacing the fuse without checking the power module, contact condition, cooling path and fault history can hide the original failure and lead to another interruption.

Reading curves without overcomplicating it

Semiconductor fuse datasheets often show time-current curves, peak let-through curves and I²t values. The purpose is not to make selection harder. The curves show whether the fuse operates fast enough at the available fault current and whether the protected device can survive the energy before interruption is complete.

For a new design, the device data and circuit fault study come first. For replacement, the existing equipment manual and original fuse reference are the starting point. Cross-references can be useful, but only when the replacement has equivalent or better voltage rating, breaking capacity, I²t behaviour, body style and thermal performance.

When the application is safety-critical, high-power or expensive to repair, selection should be verified against the manufacturer data and the equipment design rather than based on a visual match.

Body style, holders and thermal path

Semiconductor fuses often use square-body, bolted, cylindrical or special high-speed formats. The mechanical form is part of the electrical performance because heat must leave the fuse body through its terminals, holder, busbar and surrounding air. A fuse that looks correct can run too hot if the holder is underrated, the contact pressure is poor or the enclosure temperature is higher than the datasheet assumption.

For bolted fuses, the contact surfaces and tightening method matter. For cylindrical high-speed fuses, the holder rating and clip condition matter. For compact converter cabinets, airflow and nearby heat-producing parts can change the practical current rating. This is why the holder and mounting system should be checked together with the fuse link.