Ripple voltage is the small residual AC fluctuation that remains on a DC power supply’s output after converting from AC. No AC-to-DC conversion is perfectly smooth, so every power supply produces some amount of this unwanted voltage variation riding on top of the steady DC level. In a well-designed power supply, ripple is typically less than 1% of the output voltage, but even small amounts can cause problems for sensitive electronics.
How Ripple Voltage Occurs
Every electronic device that plugs into a wall outlet needs to convert the alternating current from the grid into the direct current its circuits require. This conversion, called rectification, uses diodes to force the AC waveform into a single polarity. But the result isn’t a flat, steady voltage. It’s a pulsating wave that rises and falls with each cycle of the original AC signal.
To smooth out those pulses, a capacitor is placed across the output. During each pulse, the capacitor charges up and stores energy. Between pulses, when the diode stops conducting, the capacitor discharges its stored energy into the load to keep the voltage from dropping to zero. This charge-and-discharge cycle is the direct physical cause of ripple. The output voltage rises slightly as the capacitor charges, then sags as it discharges, creating a sawtooth-like variation on top of the DC level.
Three factors control how large that variation is: the size of the capacitor, the amount of current the load draws, and the frequency of the pulses. A larger capacitor stores more energy and discharges more slowly, producing less ripple. A heavier load drains the capacitor faster, increasing ripple. And higher-frequency pulses give the capacitor less time to discharge between refills, which also reduces ripple.
Ripple in Switching vs. Linear Supplies
The character of ripple depends heavily on the type of power supply. In a traditional linear supply, ripple occurs at the rectified mains frequency: 100 Hz in countries with 50 Hz mains, or 120 Hz in 60 Hz countries. This is a relatively low-frequency ripple that’s straightforward to filter with large capacitors.
Switching power supplies work differently. They chop the DC voltage at a much higher frequency, typically between 100 kHz and 2 MHz, then use an inductor and capacitor to smooth the output. The ripple in these supplies occurs at the switching frequency rather than the mains frequency. A modern switching regulator might produce just 4 mV of peak-to-peak ripple under normal conditions. The tradeoff is that this high-frequency ripple can be harder to filter completely and may inject noise into sensitive analog circuits. That’s why switching regulators are sometimes replaced with low-dropout regulators (LDOs) in noise-sensitive applications, trading energy efficiency for a cleaner output.
How Ripple Is Measured
Ripple is most commonly expressed as a peak-to-peak voltage: the difference between the highest and lowest points of the fluctuation. But engineers also use a metric called the ripple factor, which gives a sense of how “clean” the DC output is relative to its average level. The ripple factor is the ratio of the AC component’s effective value to the DC component:
Ripple factor = Vac / Vdc
where Vac is calculated from the total output voltage and the average DC voltage. A lower ripple factor means a smoother, more stable DC output. A perfect DC supply would have a ripple factor of zero.
To physically measure ripple, you connect an oscilloscope probe across the power supply’s output terminals, as close to the output capacitor as possible. The key setting is AC coupling, which strips away the large DC voltage and lets you see only the small AC ripple riding on top. You’ll also want to limit the oscilloscope bandwidth to around 20 MHz to filter out high-frequency noise that isn’t actually part of the ripple. A differential probe can improve accuracy when common-mode noise is present. Proper grounding of the probe is critical, since a poor ground connection can introduce artifacts that look like ripple but aren’t.
Acceptable Ripple Levels
What counts as “acceptable” depends on the application. The ATX specification for computer power supplies sets the limit at 120 mV peak-to-peak on the 12V and negative 12V rails, and 50 mV on the 5V, 3.3V, and standby rails. In practice, modern PC power supplies perform well below these limits.
Audio equipment is particularly sensitive to ripple because any AC variation on the power rail can couple into the signal path as an audible hum. Precision analog circuits, medical instruments, and laboratory equipment often demand ripple in the single-digit millivolt range or lower. Digital logic, by contrast, tolerates ripple more easily as long as it stays within the voltage margins of the chips being powered.
How Ripple Affects Components
Ripple doesn’t just degrade signal quality. It physically wears out components, particularly the electrolytic capacitors used for filtering. When ripple current flows through a capacitor, the capacitor’s internal resistance converts some of that current into heat. Electrolytic capacitors are especially vulnerable because they have higher internal losses compared to ceramic or film capacitors.
This internal heating directly shortens capacitor lifespan. According to Nippon Chemi-Con, a major capacitor manufacturer, the allowable temperature rise from ripple current depends on the ambient temperature. At 85°C ambient, a capacitor can tolerate a 15°C internal temperature rise from ripple. At 105°C ambient, the margin shrinks to just 5°C. Exceed these limits and the capacitor’s life expectancy drops sharply, following an exponential relationship where each 5°C rise from ripple current roughly halves the expected lifetime.
The ripple current rating on a capacitor’s datasheet is specified at a particular frequency, usually 120 Hz or 100 kHz. Since internal resistance changes with frequency, a capacitor that handles ripple well at one frequency may overheat at another. This is a common design pitfall when capacitors are used in switching supplies where the ripple frequency differs from the rated test frequency.
Techniques for Reducing Ripple
The most basic approach is increasing the filter capacitance. A larger capacitor discharges more slowly, reducing the voltage sag between charging pulses. But beyond a certain point, simply adding capacitance runs into diminishing returns, physical size constraints, and cost.
For switching supplies, increasing the output inductor value reduces ripple by more aggressively smoothing the current waveform before it reaches the output capacitor. This is one of the simplest design-level adjustments available.
A popular technique is adding a second-stage LC filter using a ferrite bead (a small magnetic component that acts as a frequency-dependent resistor) paired with an additional capacitor. This creates a low-pass filter that attenuates the switching frequency ripple before it reaches the load. Using multiple smaller ceramic capacitors in parallel rather than a single large one can also help, because the parallel combination has lower parasitic inductance and maintains low impedance across a wider frequency range.
For the cleanest possible output, an LDO regulator can be placed after the switching converter. The LDO acts as an active filter, using its internal feedback loop to reject ripple at the switching frequency. This is the standard approach when powering sensitive analog circuits like precision data converters or RF components from a switching supply. Feedthrough capacitors, which are three-terminal ceramic capacitors with very low parasitic inductance, offer another option for applications that need low impedance across a broad frequency range.
In practice, most designs use a combination of these methods. A switching converter provides efficient bulk power conversion, followed by passive LC filtering and sometimes an LDO for the final stage of cleanup, depending on how much ripple the downstream circuit can tolerate.