Influence of Power Measurement Settings

Rigorous testing is very important when designing power supplies. Hardware measurements are essential to accomplish this task. Of course, many errors may also creep in during this measurement process. In this short article on power management tips, we will examine the effects of the connecting wires between the power supply to be tested and the load. When making a quick connection to a circuit board in the lab, the setup is usually as shown in Figure 1. In this case, a long connecting wire connects the power supply to be tested to the electronic load (shown on the right side of the figure). The two wires are placed randomly on the lab bench with a large circuit area.

Figure 1. Connection wires randomly placed between the power board and the loads

For a neater setup, see Figure 2. In this case, the two wires are twisted around each other to minimise the loop area in the circuit. In theory, this will reduce the parasitic inductance of the connecting wires between the power supply to be tested and the load. This is as far as theory is concerned. So how exactly do the different setups in Figures 1 and 2 affect the measurements? To see this, in the two different setups shown in Figures 1 and 2, we connected an ADP2386 evaluation board as a buck converter with an output current of up to 6 A. We then measured the response of the output voltage to load transients. This will show the practical effect of optimising the location of the connecting wires. In this example, the ADP2386 converts a 5 V supply voltage to a 3.3 V output voltage. A load transient is generated by the electronic load, which changes from 10 mA to 4 A in about 30 μs. In both of these cases, the length of the connection line is 1 metre.

Figure 2. Carefully placed connecting wires between the power supply board and the loads

Figure 3. Output voltage load transient response with connecting wires randomly placed (as shown in Fig.1)

Figure 3 shows voltage spikes in the AC-coupled output voltage during the load transient response with the measurement setup shown in Figure 1. The peak voltage is approximately 103 mV. In contrast, Fig. 4 shows the measurement results when the leads of the connecting line are neatly twisted together (as shown in Fig. 2). At this point, the voltage spike in the output voltage is only about 96 mV. this corresponds to a difference of about 7 mV, which suggests that the neat arrangement of the connecting wires can lead to an improvement of about 7 per cent in the transient test results for this load.

Figure 4. Output voltage load transient response with carefully placed connection lines (as shown in Fig. 2)

It is therefore clear to see that a good measurement setup provides more accurate results. In addition to the geometrical placement of the connecting wires between the power supply to be tested and the load, the length of the cable and the corresponding type of connection (in this case alligator clips or solder joints) are also important. The shorter the leads, the lower the parasitic inductance and the lower the influence on the load transient test results. The shortest possible connecting wire should always be used.

The results show that neatly twisted together leads will definitely have an effect on the measurement results, so it is worth the extra effort to twist the leads for the measurement setup.