Inrush Current and Cold Start: Why a 100 W Driver Can Trip a 20 A Breaker

Every switching power supply draws a surge of current the instant it is energised, and the size of that surge has nothing to do with the driver's steady-state wattage. A 100 W driver that runs on under 1 A can momentarily pull tens of amps for a few milliseconds at power-on. That spike is the inrush current, and it is the reason a bank of drivers on one circuit can nuisance-trip a breaker that the nameplate current says should be fine.

The cause is the input capacitor. When mains is first applied, the input filter capacitors are completely discharged and look like a dead short. The rectifier charges them as fast as the line can deliver, producing a current spike that is limited only by the source impedance and the wiring resistance. The peak can be several times the steady-state current, and for a large driver feeding many strips it can reach tens of amps for a fraction of a cycle.

The common cure is a negative temperature coefficient (NTC) thermistor in series with the input. At power-on the NTC is cold and has a high resistance, which slows the charging of the capacitor and limits the surge. As current flows through it, resistive heating drops its resistance to a small fraction of its cold value, so under normal operation it introduces almost no loss. It is a simple, passive, self-correcting limiter, and it is why most good drivers list a measured inrush current rather than letting the raw spike speak for itself.

There are limits to what an NTC can do. After a hot restart, the thermistor is still warm and low-resistance, so a second energisation within a few seconds may see a larger inrush than the first. Active soft-start circuits, which use a controlled switching element and a gradual ramp, avoid this and deliver a repeatable inrush on every start, but they add cost. For a driver that powers up once and runs, the NTC is sufficient; for frequent switching, an active limiter is worth specifying.

The practical problem appears when many drivers share one circuit. Ten drivers, each with a modest inrush of 20 A, do not start simultaneously by accident, but if they are switched on together they can draw a combined transient that trips the breaker even though the running current is well within rating. The fix is to stagger the power-up, either with individual switches, a timed sequence, or a breaker with a slow-trip characteristic. The datasheet's inrush figure and the measured cold-start surge are the numbers to add up when sizing a lighting subpanel.

Inrush is also regulated. The relevant limits sit under the IEC 61000-3-3 standard, which restricts voltage fluctuations caused by equipment on the mains. For very large drivers or many-driver installations, the cumulative inrush is part of the EMC assessment. A driver whose datasheet lists the inrush current, the pulse duration, and the measurement conditions is giving the designer what is needed to size the circuit correctly; a datasheet that omits inrush leaves the breaker-tripping surprise for the field.

There is a subtle interaction between inrush and dimming that installers should know. A triac dimmer needs a minimum holding current to stay latched, and at the moment of power-on the driver's inrush is actually helpful for that, because it pulls enough current to fire and hold the triac. The problem appears on the reverse side: a driver with an aggressive soft-start that keeps the inrush very low may not provide enough current to latch some dimmers reliably, causing a click or a refusal to start. This is why a driver's compatibility list and its startup behaviour are tested together, and why a driver that starts cleanly on the bench may still need the dimmer checked on the wall.