EMC and Surges: Why a Quiet, Protected Driver Is a Reliable Driver

Electromagnetic compatibility is the least glamorous subject in lighting and one of the most practical. A driver is an EMC event in its own right: it rectifies the mains, switches at tens of kilohertz and modulates a power stage, and all of that activity can leak noise back onto the supply or radiate into the room. How well the manufacturer contains that noise is a reliability question, not a paperwork one.

The relevant standards form a well-defined set for lighting equipment in Europe, where the CE mark depends on them. EN IEC 55015 sets the limits for conducted and radiated emissions from lighting products. EN IEC 61547 covers immunity — the product's ability to keep working when the environment pushes back. IEC 61000-3-2 limits the harmonic currents a lighting product may draw from the mains, the same distortion story as power factor. A driver that passes these tests has demonstrated, under defined laboratory conditions, that it neither pollutes the supply it shares with other equipment nor misbehaves when exposed to interference. That is not a marketing claim; it is a measured outcome. The same set of standards appears in equivalent form in most other markets, so the discipline is global even where the mark on the label differs.

The interference a driver faces in the field is mostly transient: switching surges from nearby equipment, and the distant effects of lightning. Surge immunity testing follows IEC 61000-4-5, which applies a standardized combination wave — a voltage pulse of 1.2/50 µs combined with a current pulse of 8/20 µs — at test levels chosen for the environment. A lighting product is expected to survive defined surge levels without damage and without latching up, and the limits of that protection are exactly why building-level surge protection devices still have a job. In practical terms, surge protection has two layers: the driver's internal protection, sized to survive defined tests, and the building's external protection, which absorbs the energy before it reaches the equipment. Neither replaces the other.

Dimming adds a special twist. The 0–10 V control pair is a long, low-voltage antenna in a hostile environment, and noise coupled into it appears directly as a wrong light level or a flicker event, because the control input has no packets to reject — it simply believes what it sees. Control wiring discipline — twisted pair, separation from mains, no shared conduits — and a driver whose control input is properly filtered are therefore part of the EMC design, not cosmetic details. It is also one more reason flicker-free claims deserve measurement on site: a control line that picks up noise can make even a well-filtered driver misbehave at the wrong moment.

Installation practice closes the loop. A Class I driver must have its protective earth connected: the earth is part of the noise-containment strategy, giving common-mode currents a defined path home. Mains connections should be short and firm, the control pair should be dressed away from power cables, and where the supply is known to be rough, a properly rated surge protection device at the distribution point is cheaper than a service visit. A quiet, protected driver is not merely a compliant driver. It is the one that still holds its dimming curve five years later. And one maintenance habit protects any dimming system for nothing: re-torque the screw terminals after the first few months of operation, because thermal cycling loosens connections that were tight on day one.