“Oscilloscopes are the most basic measurement equipment for today’s engineers to test and characterize power supply designs. Keysight’s InfiniiVision X series is the only oscilloscope in the industry that can automatically measure loop response. Compared with Keysight’s network analyzer test results, the gain and phase test results are very close.
“
The oscilloscope is the most important power supply test and characterization tool. At present, many oscilloscopes (including Keysight’s InfiniiVision X-series oscilloscopes) can provide dedicated power measurement options to help engineers automatically complete many important tests. Figure 1 shows the power measurement items supported by Keysight’s InfiniiVision 3000T, 4000, and 6000 X-series oscilloscopes using the power measurement options (DSOX3PWR, DSOX4PWR, DSOX6PWR). Frequency response measurement is a unique measurement function of Keysight, which includes control loop response (Bode diagram) and power supply rejection ratio (PSRR). Such specific stimulus response measurement is usually done by a low-frequency network analyzer. However, since Keysight’s InfiniiVision X-series oscilloscope has a built-in function/arbitrary waveform generator, it can also be used for this type of measurement.
Figure 1: Power characterization measurement items supported by Keysight InfiniiVision X-series oscilloscopes
Control loop response (Bode diagram)
The power supply is actually an amplifier that includes a negative feedback control loop, as shown in Figure 2. This means that although you can think of the power supply as a DC amplifier, it actually amplifies the AC signal and responds to changes in output conditions, such as load changes.
In order to implement the control loop response test, you need to inject an error signal into the feedback path of the control loop. In Figure 2, this feedback path refers to the resistor divider network of R1 and R2. We need to connect a smaller resistor in series with the feedback loop to inject an error signal. The 5 Ω injection resistance shown in the schematic is insignificant compared to the series impedance of R1 and R2. Therefore, you may consider using this low-resistance injection resistor (Rinj) as a test device for long-term use. In addition, an injection transformer (such as Picotest J2101A) is used to isolate the AC interference signal, so as not to generate any DC offset.
The measurement system here is an InfiniiVision X-series oscilloscope with a built-in WaveGen function/arbitrary waveform generator, which can measure the AC voltage level on the top of the feedback network (Vin) and the DC output (Vout) of the power supply. The oscilloscope can also calculate the gain at each test frequency point in the sweep frequency band, dB=20Log(Vout/Vin). At the same time, the phase difference between Vin and Vout is also measured.
Figure 2: Power closed-loop feedback network and oscilloscope connection settings during loop response test
Input and output detection
Rigorous and accurate testing technology is a prerequisite for loop response testing. The ripple voltage of Vin and Vout at certain test frequency points is very small, and it is easy to be overwhelmed by the noise floor of the oscilloscope or the noise of the power supply under test. Increasing the signal-to-noise ratio of the test can increase the dynamic range of the frequency response test. Although most oscilloscopes are equipped with 10:1 passive probes as standard, the use of 1:1 passive probes can better reduce the impact of the oscilloscope’s noise floor and power switch noise, thereby improving the test signal-to-noise ratio. Keysight recommends using the 1:1 passive probe N2780A with 35MHz bandwidth for loop response testing.
Shortening the ground wire length of the probe is also very helpful for this test. The standard ground wire of the probe sometimes absorbs switching noise like an antenna. If there are binding posts near the Vin and Vout test points, you can remove the standard long ground wire and probe cap, and then directly connect the probe as shown in Figure 3. The probe at the front end and the ground leaned against the terminal. If there is no binding post, find a closer ground point on the PCB, and then use the spring ground wire accessory attached to the passive probe to make a shorter ground wire connection, or reserve a probe interface on the PCB board during design. Get the best low-noise ground connection without the need to fix the probe by hand.
Figure 3: Using a short ground wire measurement can provide the best signal-to-noise ratio in the loop response test
Perform loop response measurement
Figure 4 shows the isolation transformer and 1:1 passive probe that a switching power supply evaluation board needs to configure for loop response testing. The isolation transformer uses Picotest’s N2101A injection transformer. This power supply test board has been designed with a 5Ω injection resistor in the feedback loop.
Figure 4: The configuration of the loop response test isolation transformer and 1:1 passive probe
The loop response test needs to use an InfiniiVision X series oscilloscope with power test option. The first step is to select the loop response test in the power test item list, and then select the Setup & Application menu as shown in Figure 5. We use Fix the default setting of 200mVpp peak-to-peak value and 50Ω impedance for this test. Since the injection resistance connected to the isolation transformer is only 5Ω, the actual voltage applied to the injection resistance is only about 36mVpp.
Figure 5: Loop response test parameter setting
Please note that most frequency response analyzers or vector network analyzers set the injection voltage to Vrms or dBm and are based on 50Ω injection resistance. If you want to compare the test results of different devices (oscilloscope, frequency response analyzer, network analyzer), Even when the voltage units are different (Vpp, Vrms, dBm), make sure that the injection voltage settings are the same.
After setting the test parameters, click the Apply button directly, as shown in Figure 6, the blue-green and orange are the gain and phase curves in the 100Hz to 20MHz scanning frequency band, respectively. The oscilloscope also automatically displays the phase margin (PM) and gain margin (GM) test results at the corner of the screen.
The loop response test needs to use an InfiniiVision X series oscilloscope with power test option. The first step is to select the loop response test in the power test item list, and then select the Setup & Application menu as shown in Figure 5. We use Fix the default setting of 200mVpp peak-to-peak value and 50Ω impedance for this test. Since the injection resistance connected to the isolation transformer is only 5Ω, the actual voltage applied to the injection resistance is only about 36mVpp.
Figure 6: The gain curve and phase curve of the loop response test at a fixed injection voltage
We can see that the gain and phase results are jittered at low frequencies in the test results. This result indicates that the signal-to-noise ratio may be too low at this frequency. A simple way to improve the signal-to-noise ratio is to increase the amplitude of the injected voltage in all frequency bands. However, increasing the injection voltage may cause signal distortion or incorrect phase margin results when the 0dB crossover occurs.
A better way to improve the signal-to-noise ratio is to use the amplitude configuration function in the measurement tool shown in Figure 7 to customize different injection voltages. Through the amplitude configuration function, you can use it in the sensitive frequency range of the power supply to be tested. Low-amplitude injection voltage, use high-amplitude injection voltage in the frequency range where the power supply under test is not sensitive. The switching power supply feedback network is generally the most sensitive near the 0dB crossing point. The power board we measured is about 60KHz.
Figure 7: Use amplitude configuration to optimize signal-to-noise ratio
Figure 8 shows the test results of the loop response gain and phase after the customized injection voltage. It can be seen that the jitter of the test result caused by the insufficient signal-to-noise ratio has obviously become very small.
Figure 8: Gain and phase test results after optimizing the injected voltage configuration
How to determine the most suitable injection voltage? Some engineers suggest to use a relatively high injection voltage for testing at the beginning, and then gradually reduce the amplitude to perform multiple tests until the 0dB crossover frequency stabilizes at a certain frequency point. The frequency response test generally repeats the test several times to obtain the best dynamic range by iterating the appropriate injection voltage. Using the frequency response test of the Keysight oscilloscope solution, you can generally find the appropriate injection voltage faster.
Please note that you can also use Keysight’s oscilloscope solution to perform phase and gain measurements at a single frequency point. This solution allows you to run the test at a single frequency point, such as a certain frequency point near the estimated 0dB crossover frequency, and then manually adjust the voltage and frequency of the injected signal in the waveform generator setting menu of the oscilloscope, and you can visually observe the real-time waveform The change.
For example, you can run the entire sweep range test with a fixed low injection voltage as shown in Figure 6 to determine the approximate 0dB crossover frequency. In this example, we know that the 0dB crossover frequency is about 62KHz, and then run the test of a single frequency point of 62KHz and gradually increase the injection voltage until the signal distortion is observed. Once you have determined the maximum injection voltage near the 0dB crossover frequency, you can simply use the injection voltage in all frequency bands. You can perform more single frequency point tests at different frequency points to optimize the injection voltage amplitude at low frequency or high frequency.
Summarize
Oscilloscopes are the most basic measurement equipment for today’s engineers to test and characterize power supply designs. Keysight’s InfiniiVision X series is the only oscilloscope in the industry that can automatically measure loop response. Compared with Keysight’s network analyzer test results, the gain and phase test results are very close.
The Links: SKM200GAR125D SKM600GA124D AUO LCD
