ADI experts analyze switching regulator outputs to effectively speed up power supply design

　　Minimizing the output ripple and transients of a switching regulator is important, especially when powering noise-sensitive devices such as high-resolution ADCs, where the output ripple will appear as a distinct spur on the ADC output spectrum. To avoid degradation in signal-to-noise ratio (SNR) and spurious-free dynamic range (SFDR) performance, switching regulators are often replaced with low-dropout regulators (LDOs), sacrificing the high efficiency of the switching regulator in exchange for a cleaner LDO output. Understanding these artifacts allows designers to successfully integrate switching regulators into more high-performance, noise-sensitive applications.

　　This article describes effective methods for measuring output ripple and switching transients in switching regulators. Measuring these parameters requires great care, as a poor setup can result in erroneous readings and parasitic inductance is introduced by the loop formed by the oscilloscope probe signal and ground leads. This increases the magnitude of transients associated with fast switching transients, so short connections, effective methods, and wide bandwidth performance are essential. Here, the ADP2114 dual 2 A/single 4 A synchronous step-down DC-DC converter is used to demonstrate methods for measuring output ripple and switching noise. This step-down regulator has high efficiency and can switch at frequencies up to 2 MHz.

　　Output ripple and switching transients

　　Output ripple and switching transients depend on the regulator topology as well as the values ​​and characteristics of the external components. Output ripple is the residual AC output voltage that is closely related to the switching operation of the regulator. Its fundamental frequency is the same as the switching frequency of the regulator. Switching transients are high-frequency oscillations that occur during the switching transitions. Their amplitude is expressed as the maximum peak-to-peak voltage, which is difficult to measure accurately because it is highly dependent on the test setup. Figure 1 shows an example of output ripple and switching transients.

　　Figure 1. Output ripple and switching transients.

　　Output Ripple Considerations

　　The inductor and output capacitor of the regulator are the main components that affect the output ripple. Smaller inductors produce faster transient response at the expense of larger current ripple, while larger inductors produce smaller current ripple at the expense of slower transient response. Using capacitors with low effective series resistance (ESR) can minimize output ripple. Ceramic capacitors with dielectric X5R or X7R are a good choice. Large capacitors are usually used to reduce output ripple, but the size and number of output capacitors are obtained at the expense of cost and PCB area.

　　Frequency domain measurements

　　For power engineers, it is very useful to consider the frequency domain when measuring unwanted output signals, which provides a better perspective on at which discrete frequencies the output ripple and its harmonics are located, and at what different power levels each corresponds. An example of a spectrum is shown in Figure 2. This type of information can help engineers determine if the selected switching regulator is suitable for their wideband RF or high-speed converter application.

　　To make frequency domain measurements, connect a 50Ω coaxial cable probe across the output capacitor. The signal passes through the DC blocking capacitor and terminates with a 50Ω termination resistor at the input of the spectrum analyzer. The DC blocking capacitor blocks DC current from passing through the spectrum analyzer, avoiding DC loading effects. The 50Ω transmission environment minimizes high-frequency reflections and standing waves.

　　The output capacitor is the main source of output ripple, so the measurement point should be as close as possible. The loop from the signal tip to the ground point should be as small as possible to minimize additional inductance that may affect the measurement results. Figure 2 shows the output ripple and harmonics in the frequency domain. The ADP2114 produces 4 mV pp output ripple at the fundamental frequency under the specified operating conditions.

　　Figure 2. Frequency domain plot using a spectrum analyzer.

　　Time domain measurements

　　When using an oscilloscope probe, avoid using a long ground lead to avoid ground loops, as the loop formed by the signal tip and the long ground lead creates additional inductance and higher switching transients.

　　When measuring low-level output ripple, use a 1× passive probe or 50Ω coaxial cable instead of a 10× oscilloscope probe because the 10× probe attenuates the signal by a factor of 10, pushing the low-level signal down to the oscilloscope noise floor. Figure 3 shows a suboptimal probing method. Figure 4 shows the waveform measurement results using a 500MHz bandwidth setting. High-frequency noise and transients are measurement artifacts caused by the loop formed by the long ground lead and are not inherent to the switching regulator.

　　Figure 3. Ground loop creates output error.

　　Figure 4. Switching node (1) and AC-coupled output waveform (2)

　　There are several ways to reduce stray inductance. One method is to remove the long ground lead of a standard oscilloscope probe and connect its barrel to a ground reference point. Figure 5 shows the tip and barrel method. However, in this case, the tip is connected to the wrong point on the regulator output instead of directly to the output capacitor, which should be the correct method. The ground lead has been removed, but the inductance caused by the trace on the PCB is still there. Figure 6 shows the waveform results when using a 500MHz bandwidth setting. Because the long ground lead has been removed, the high-frequency noise is reduced.

　　Figure 5. Probing at any point on the switch output using the tip and tube method

　　Figure 6. Switching node (1) and AC-coupled output waveform (2)

　　As shown in Figure 7, using a grounded coil to directly probe the output capacitor produces nearly optimal output ripple. The noise of the switching transients is improved, and the trace inductance on the PCB is greatly reduced. However, the low-amplitude signal profile is still clearly superimposed on the ripple, as shown in Figure 8.

　　Figure 7. Tip and barrel probing on output capacitor through grounded coil.

　　Figure 8. Switching node (1) and AC-coupled output waveform (2)

　　Best Practices

　　The best way to probe the output of a switch is to use a 50Ω coaxial cable that is maintained in a 50Ω environment and terminated with a selectable 50Ω oscilloscope input impedance. Placing a capacitor between the regulator output capacitor and the oscilloscope input blocks the flow of DC current. The other end of the cable can be soldered directly to the output capacitor with very short flying leads as shown in Figure 9 and Figure 10. This maintains signal integrity when measuring very low level signals over a wide bandwidth. Figure 11 shows a comparison of the tip and barrel method and the 50Ω coaxial method at the output capacitor for probing at a 500 MHz measurement bandwidth.

　　Figure 9. Optimal probing method using terminated 50Ω coaxial cable.

　　Figure 10. Example of optimal detection method

　　Figure 11. Switch node (1), tip and tube method (3), 50 Ω coaxial method (2)

　　Comparing these methods shows that using coaxial cable in a 50Ω environment produces more accurate results with less noise, even at a 500 MHz bandwidth setting. Changing the oscilloscope bandwidth to 20 MHz removes the high frequency noise, as shown in Figure 12. The ADP2114 produces 3.9 mV pp output ripple in the time domain, which is close to the 4 mV pp value measured in the frequency domain using the 20 MHz bandwidth setting.

　　Figure 12. Switching node (1) and output ripple (2)

　　Measuring switching transients

　　Switching transients have lower energy but higher frequency content than the output ripple. This occurs during the switching transition and is usually normalized to include the peak-to-peak value of the ripple. Figure 13 shows the switching transient measurement results using a standard oscilloscope probe with a long ground lead compared to using a 50Ω coaxial terminated cable (500 MHz bandwidth). Often, the ground loop created by the long ground lead will produce higher than expected switching transients.

　　Figure 13. Switch node (1), standard oscilloscope probe (3), 50Ω coaxial termination (2)

　　in conclusion

　　Output ripple and switching transient measurement methods are very important considerations when designing and optimizing system power supplies for low noise, high performance converters. These measurement methods provide accurate and reproducible results in both the time and frequency domains. Maintaining a 50Ω environment is very important when measuring low level signals over a wide frequency range. A simple, low-cost method for making this measurement is to use a properly terminated 50Ω coaxial cable. This method can be used for a variety of switching regulator topologies.