Output Ripple: Where It Comes From and How It Reaches the LEDs
A regulated output is never perfectly flat. The voltage at the output terminals carries a small alternating component, the ripple, which comes from the switching action and from the rectified mains. Ripple is specified in millivolts peak-to-peak or as a percentage of the output, measured at a stated load and input voltage with a defined bandwidth. On an LED system the ripple matters because the light follows the current, and the current follows the voltage.
Two frequency bands dominate. The low-frequency component appears at twice the mains frequency, 100 Hz on a 50 Hz supply, because the rectified waveform delivers energy in pulses and the bus capacitor smooths only part of it. The high-frequency component comes from the switching converter, typically tens to hundreds of kilohertz, and is attenuated by the output filter but not eliminated. The low-frequency part is harder to remove, because suppressing it requires a larger bus capacitor or a faster control loop.
The ripple reaches the LED through the load. In a constant-voltage system with a series resistor, the current ripple is the voltage ripple divided by the resistance, so a 240 millivolt ripple on a 24 volt bus with a 5 ohm resistor produces about 48 milliamps of current ripple, which is a large fraction of a one amp load. A constant-current module regulates the current directly and rejects most of the bus ripple, which is one reason module-level regulation is used in quality-sensitive applications.
The visible consequence is modulation. Light output tracks current, so ripple becomes a periodic change in brightness. At 100 Hz the eye does not see separate flashes but may perceive flicker in peripheral vision or during movement, and a camera can record banding. The depth of modulation, not only the average level, determines whether the effect is objectionable, which is why flicker metrics are stated separately from the DC output accuracy.
Measuring ripple correctly takes care. Use a probe with a short ground lead, or a coaxial connection, because a long ground lead picks up switching noise and shows ripple that is not there. Limit the scope bandwidth to twenty megahertz to avoid counting radio-frequency noise as ripple, and measure at the specified load and input voltage. Comparing a measurement taken with a long clip lead against the datasheet value is a common source of false complaints.
The datasheet should state the ripple in millivolts peak-to-peak together with the test conditions. For the PV100, the 24 volt output is filtered to a level that keeps the LED current stable under normal loads, and the value should be read together with the load and the wiring. In installations with long output cables, the cable inductance and the load capacitance can change the ripple at the fixture, so the measurement at the driver terminals is the reference.
Temperature changes the picture as well. Electrolytic capacitors lose capacitance and gain equivalent series resistance as they age and cool, so a driver that meets its ripple specification when new may exceed it after years of service. Design margin, not the initial measurement, is what keeps the ripple stable over the life of the installation.
The load also matters. Ripple measured at full load can be lower than at light load, because the control loop enters a different operating mode. A datasheet value taken at one operating point should not be assumed to hold across the dimming range, and a specification that matters at low output should be verified at low output.
Cold Start: What Happens Inside a Driver Below Zero
Dielectric Strength and Hipot Testing: What the Isolation Rating Means
Related Article