Earth Leakage Current: Why a Bank of LED Drivers Trips the RCD
Every mains-powered driver contains capacitors connected between the mains and the protective earth, and those capacitors leak a small AC current. One driver leaks a fraction of a milliamp, which is harmless. Forty drivers on the same circuit leak forty times that, and the total can reach the threshold where a residual current device decides there is a fault and opens the circuit. The result is an installation that trips the earth leakage protection at switch-on and works perfectly when tested.
The leakage path is the Y capacitor network. Y capacitors are placed across the isolation barrier to shunt common-mode noise to earth, and their impedance at 50 Hz is not infinite. At 230 volts a capacitor of a few nanofarads passes a current of roughly a milliamp, and the exact value depends on the capacitor size and the number of stages. The current is capacitive, in quadrature with the voltage, but the residual current device responds to the magnitude.
Standards limit the leakage of a single product. IEC 60598-1 sets touch current limits for luminaires, and a compliant driver stays below them by design, because the limits are written per product and assume the product is used alone. The installation standard is where the total appears, and the person who signs the design is responsible for the sum, not for the individual driver. This split between product compliance and installation compliance is where the confusion starts.
The type of residual current device matters. A Type AC device responds to sinusoidal AC residual current only. A Type A device also responds to pulsating DC, which can appear when a load rectifies the mains, and a Type B device responds to smooth DC as well. Modern drivers with power factor correction have rectifiers and switching stages, and a Type A device is the usual minimum recommendation for lighting circuits.
The arithmetic on site is simple. If a device is rated at 30 milliamps and each driver leaks 0.75 milliamps, forty drivers reach the threshold before any fault current exists. Add the inrush pulse, which contains a capacitive component, and the device can trip on energisation even though nothing is wrong. Splitting the circuit across two devices, or using a device with a higher threshold where the standard permits, restores the margin.
Measuring the leakage is not difficult. A leakage clamp meter placed around the earth conductor of the circuit, with all drivers energized at full output, reads the total. Comparing that with the device rating shows how much margin remains, and repeating the measurement at minimum dimming shows whether the leakage changes with the operating point. A reading close to the threshold explains intermittent trips that no insulation test can reproduce.
The design response is to plan for the sum. Specify the leakage per driver, count the drivers per circuit, and choose the residual current device and the number of circuits from that total. Where a large installation needs many drivers on one feed, a higher-threshold device may be permitted for the application, or the circuits can be divided. The decision belongs in the electrical design, not in the driver datasheet.
For the PV100, the leakage is part of the product EMC design and is covered by the safety certification, but the installation total is not. A project using forty or fifty units on one residual current device should calculate the sum before the first switch-on. The calculation takes minutes, and it prevents the callout that follows a trip on the morning of handover.
Flammability of Driver Enclosures: UL 94 and the Glow-Wire Tests
Wireless 0-10V Adapters: Networking an Installed Base
Related Article