SNVS124F November 1999 – April 2021 LM2596
The output voltage of a switching power supply operating in the continuous mode will contain a sawtooth ripple voltage at the switcher frequency, and may also contain short voltage spikes at the peaks of the sawtooth waveform.
The output ripple voltage is a function of the inductor sawtooth ripple current and the ESR of the output capacitor. A typical output ripple voltage can range from approximately 0.5% to 3% of the output voltage. To obtain low ripple voltage, the ESR of the output capacitor must be low; however, exercise caution when using extremely low ESR capacitors because they can affect the loop stability, resulting in oscillation problems. TI recommends a post ripple filter if very low output ripple voltage is required (less than 20 mV) (see Figure 9-10). The inductance required is typically between 1 μH and 5 μH, with low DC resistance, to maintain good load regulation. A low ESR output filter capacitor is also required to assure good dynamic load response and ripple reduction. The ESR of this capacitor may be as low as desired, because it is out of the regulator feedback loop. Figure 8-9 shows a typical output ripple voltage, with and without a post ripple filter.
When observing output ripple with a scope, it is essential that a short, low inductance scope probe ground connection be used. Most scope probe manufacturers provide a special probe terminator which is soldered onto the regulator board, preferably at the output capacitor. This provides a very short scope ground, thus eliminating the problems associated with the 3-inch ground lead normally provided with the probe, and provides a much cleaner and more accurate picture of the ripple voltage waveform.
The voltage spikes are caused by the fast switching action of the output switch and the diode, the parasitic inductance of the output filter capacitor, and its associated wiring. To minimize these voltage spikes, the output capacitor should be designed for switching regulator applications, and the lead lengths must be kept very short. Wiring inductance, stray capacitance, as well as the scope probe used to evaluate these transients, all contribute to the amplitude of these spikes.
When a switching regulator is operating in the continuous mode, the inductor current waveform ranges from a triangular to a sawtooth type of waveform (depending on the input voltage). For a given input and output voltage, the peak-to-peak amplitude of this inductor current waveform remains constant. As the load current increases or decreases, the entire sawtooth current waveform also rises and falls. The average value (or the center) of this current waveform is equal to the DC load current.
If the load current drops to a low enough level, the bottom of the sawtooth current waveform reaches zero, and the switcher smoothly changes from a continuous to a discontinuous mode of operation. Most switcher designs (regardless of how large the inductor value is) is forced to run discontinuous if the output is lightly loaded. This is a perfectly acceptable mode of operation.
In a switching regulator design, knowing the value of the peak-to-peak inductor ripple current (ΔIIND) can be useful for determining a number of other circuit parameters. Parameters such as peak inductor or peak switch current, minimum load current before the circuit becomes discontinuous, output ripple voltage, and output capacitor ESR can all be calculated from the peak-to-peak ΔIIND. When the inductor nomographs in Figure 9-5 through Figure 9-8 are used to select an inductor value, the peak-to-peak inductor ripple current can immediately be determined. Figure 9-9 shows the range of (ΔIIND) that can be expected for different load currents. Figure 9-9 also shows how the peak-to-peak inductor ripple current (ΔIIND) changes as you go from the lower border to the upper border (for a given load current) within an inductance region. The upper border represents a higher input voltage, while the lower border represents a lower input voltage.
These curves are only correct for continuous mode operation, and only if the inductor selection guides are used to select the inductor value.
Consider the following example:
VOUT = 5 V, maximum load current of 2.5 A
VIN = 12 V, nominal, varying between 10 V and 16 V.
The selection guide in Figure 9-6 shows that the vertical line for a 2.5-A load current and the horizontal line for the 12-V input voltage intersect approximately midway between the upper and lower borders of the 33-μH inductance region. A 33-μH inductor allows a peak-to-peak inductor current (ΔIIND), which is a percentage of the maximum load current, to flow. In Figure 9-9, follow the 2.5-A line approximately midway into the inductance region, and read the peak-to-peak inductor ripple current (ΔIIND) on the left hand axis (approximately 620 mAp-p).
As the input voltage increases to 16 V, approaching the upper border of the inductance region, the inductor ripple current increases. Figure 9-9 shows that for a load current of 2.5 A, the peak-to-peak inductor ripple current (ΔIIND) is 620 mA with 12 VIN, and can range from 740 mA at the upper border (16 VIN) to 500 mA at the lower border (10 VIN).
Once the ΔIIND value is known, use these equations to calculate additional information about the switching regulator circuit.