Oscilloscope Basics - 10 Power Integrity Measurement Tips

Figure 10. Measurement results of 3.3 V power supply and 10 MHz clock in the target system. By using FFT we can verify whether the 10 MHz clock generates noise on the 3.3 V power supply.

Figure 11. Triggering on a 10 MHz clock and enabling averaging removes all random noise and signals not related to the clock. The result is the power supply noise associated with the 10 MHz clock.

Tip 9. Have enough bandwidth

In Tip 2, we discussed limiting the measurement bandwidth to only what is necessary to minimize noise for the task at hand. Users who do not have enough bandwidth to perform their tasks can fall into a similar but opposite trap—they risk not being able to detect high-frequency noise and transients, which can also negatively impact clocks and data in their systems. High-frequency power supply noise, which can be generated by switching currents from clocks and data, can also affect these devices. In many modern systems, you may need bandwidths of more than 1 GHz to see this noise, so it is important to select a probe with sufficient bandwidth. Figure 12 shows a comparison of the results of measuring 1.5 V DDR3 memory power supply noise using a common 35 MHz 1:1 probe and a 2 GHz N7020A probe. The graph shows that the wider bandwidth probe is able to more accurately capture high-frequency noise. This noise is difficult to see in many modern digital systems.

Figure 12. Comparison of 1.5 V DDR3 memory supply noise measurements using a 35 MHz 1:1 passive probe and a 2 GHz N7020A power supply probe. The lower bandwidth probe misses the high frequency noise and spikes that affect high speed digital systems.

Tip 10. N7020A Power Supply Probe

The tips mentioned previously can help you minimize the noise in your oscilloscope measurement system and find the sources of power supply noise when measuring power supply noise. These techniques can be even more useful when used in conjunction with tools specifically designed to measure power supply noise. For example, the N7020A power supply probe shown in Figure 13 (and used in some of the previous examples) is the first probe designed specifically for measuring DC power supply noise. The probe has a 1:1 attenuation ratio (Tip 3), ± 24 V offset (Tip 4), and connects to a 50 Ω oscilloscope input (Tip 1). It has a 2 GHz bandwidth to capture high-frequency noise and transients that can cause clock and data jitter (Tip 7). When used with an oscilloscope like the Keysight Infiniium S-Series, it can limit the bandwidth to a necessary level (Tip 2) to reduce noise when the full 2 ​​GHz bandwidth is not needed.

Figure 13. The N7020A power supply probe (right) is designed for measuring power supply noise and can be used with a variety of Keysight oscilloscopes, including the S-Series HD oscilloscopes (left).

Specifications and Characteristics: N7020A Power Supply Probe