Issue link: https://iconnect007.uberflip.com/i/726276
44 The PCB Design Magazine • September 2016 put terminals we have only the output capaci- tor. If we do this, we may need to apply a DC bias we have during normal operation, so that we measure the part with the correct DC oper- ating point. To remove any potential contribu- tion from the evaluation board, we can desolder the capacitor and measure it in a fixture. Ulti- mately we can obtain samples of the same ca- pacitor model and measure several of them to see how different or how identical their data looks. For example, in Figure 6 the combined plot of the evaluation board is shown with and without input power. We can see that around 0.5MHz, where the output capacitor's ESR mat- ters, the agreement is quite good, namely the impedance in that frequency range even with the converter running is primarily dictated by the output capacitor. From all the above we can conclude that the ESR of the output capacitor is around 2 to 2.5 mOhms, assuming that the plane resistance on the evaluation board be- tween the location of the output capacitor and the point where we measure the output im- pedance is in the range of one to one and half mOhms. Finally, we can confirm this by mea- suring the part in a fixture with the fixture de- embedded. Once we feel comfortable with the basic re- sults, we can look further and ask again: What did we expect or what are we looking for in the test results? One qualitative feature that we may want to see, especially if we want to use the measured impedance to create simple equivalent circuits for simulations, is the con- sistency of the impedance plots across the dif- ferent input parameters, such as load current, input voltage and possibly ambient tempera- ture. Output voltage should also be considered for the list if we plan on using the same device with different output voltage settings. To get a sense of how consistent is the small-signal impedance performance of the converter, Fig- ure 7 shows the impedance surfaces at three different input voltages. We can see that the primary variation is along the frequency axis, EVALUATING EVALUATION BOARDS Figure 6: Output impedance of the evaluation board with no input power (green trace) and with input power (blue trace) applied.