LEDs are current-driven devices: their brightness follows current, and their forward voltage shifts with temperature, which means a voltage source lets current run away as the LED heats up. Constant-current drivers solve this by regulating current directly. But switching drivers do not deliver perfectly smooth current; they deliver current with ripple, a sawtooth riding on the DC level. Uncontrolled ripple causes visible flicker, color shift, audible buzz, and shortened LED life. Controlling it is a core lighting design skill.
Where ripple comes from
Switching LED drivers regulate by rapidly switching an inductor's current, and the inductor current inherently ramps up and down each switching cycle. That ramp is the ripple. Its magnitude depends on the inductance, the switching frequency, and the input-to-output voltage differential: smaller inductors, lower frequencies, and larger voltage steps all increase ripple. The driver's control loop sets the average current; the power stage physics sets the ripple around it.
Some ripple is unavoidable in a switching design. The question is always how much the application tolerates. A shop high-bay can accept ripple that would be unacceptable in video lighting or a reading lamp, because the failure modes, flicker perception and camera artifacts, depend on context.
Why ripple matters to the LED
LED light output tracks instantaneous current, not average, so ripple modulates brightness at the switching frequency and its harmonics. At low frequencies this reads as visible flicker; at higher frequencies it is invisible to the eye but still shows up as banding in camera footage and can cause headaches or eye strain in sensitive people. Large ripple also pushes peak current above the average, stressing the LED's thermal design at the peaks even when the average looks safe.
Color shifts too. Many white LEDs shift slightly in chromaticity with current, so ripple paints a subtle color modulation onto the light. In color-critical applications like photography or retail display, that modulation is a defect even when flicker is invisible.
Design controls: inductance and frequency
The primary ripple controls are inductor value and switching frequency. Larger inductance smooths the current ramp but costs size, expense, and transient response; higher switching frequency shrinks the ripple for a given inductor but increases switching losses and EMI. The design balances these against efficiency, size, and cost targets.
Output capacitance is the second line of defense. A capacitor across the LED string shunts the AC ripple component, leaving the LED with cleaner DC. Ceramic capacitors handle the high-frequency content; their value and ESR determine how much ripple reaches the LED. Size the output capacitance for the ripple target, not just for stability, and verify under the full input voltage and temperature range, since ripple is worst at the extremes.
Layout and measurement
Layout affects ripple more than many designers expect. Long inductive loops in the switching path ring and add high-frequency content to the output; tight loops, a solid ground plane, and the output capacitor placed at the LED terminals keep the delivered current clean. Keep the feedback sense traces short and away from the switching node, since noise injected into the sense signal modulates the regulated current directly.
Measure ripple at the LED terminals with a scope, using a current probe or a small sense resistor, across the operating range. Measure at minimum and maximum input voltage and at temperature extremes if the application demands it. A design that is clean at room temperature on the bench can ripple badly at cold startup or hot soak, when inductor and capacitor values drift.
Flicker standards and system thinking
Where flicker matters, design to a metric, not to "looks fine." Percent flicker and flicker index quantify the modulation, and application standards set limits for workplaces, cameras, and health-sensitive environments. Know the target before choosing the ripple budget; it is far cheaper than discovering the problem after installation.
Think system-wide. Dimming interacts with ripple: PWM dimming at low duty cycles can alias with the switching frequency and produce visible beat flicker even when full-brightness ripple was acceptable. Mains-frequency ripple from a poorly filtered front end adds a slow modulation no output capacitor fully removes. Specify the driver, the dimmer, and the filtering as one system, and test them together under dimming, not just at full output.
Ripple control is inductor, frequency, capacitance, and layout, verified by measurement across the operating range and judged against a flicker metric the application actually needs. LEDs deserve clean current; give it to them and they repay it with stable color, silent operation, and long life.
Related Guides
- Constant Voltage Driver Voltage Drop Calculations: Low-voltage lighting runs, LED strips, landscape circuits, and cabinet lights all share a quiet failure mode: the far end of the run glows dimmer and
- Current-Limiting Resistor Calculations for Arrays: LED arrays, multiple light-emitting diodes wired together for task lighting, accent lighting, or indicators, need current limiting: LEDs are
- Low Voltage Drop Sizing on Long LED Tape Runs: LED tape looks simple until the far end of a long run glows dimmer and yellower than the start. That fade is voltage drop: the thin copper traces in
- Selecting Wire Gauges to Prevent Low Voltage DC Drops: Low-voltage DC circuits, 12V and 24V lighting, automotive accessories, solar setups, are brutally sensitive to voltage drop in a way that