[Repost] How to measure power supply ripple

皇华Ameya360

[Repost] How to measure power supply ripple [Copy link]

For hardware engineers, the use of some tools is also very important. Improper use of tools often leads to incorrect results, which may give hardware engineers a wrong judgment. This chapter will list many practical methods and techniques for using some tools that hardware engineers must know.

Measuring Power Supply Ripple

How do you measure the power supply ripple? How big is the power supply ripple you designed? How do you reduce the ripple and noise of the power supply?

The above series of questions are summarized by the author during the interview process, which are also questions that many hardware engineers like to ask during interviews. Previous blogs have more or less mentioned the filtering methods of power supply ripple and noise, so I will not repeat them here.

This section will talk about how to measure power supply ripple and noise.

Figure 1 Schematic diagram of power supply ripple and noise

The above figure is a schematic diagram of power supply ripple and noise. Ripple is the AC part of the power supply output DC, that is, the low-frequency part, and noise is the high-frequency part. As long as the power supply we design is within a reasonable range, it is generally considered to be acceptable. For example, 5V and 3.3V are considered normal below 50mV. Generally, the ripple cannot exceed 1% of the supply voltage value. Figure 2 Schematic diagram of power supply ripple using the oscilloscope side. 51)]The above figure is a schematic diagram of using an oscilloscope to measure power supply ripple and noise. For the oscilloscope we use, the probe is generally a passive probe. Regarding the selection of oscilloscope probes, there is no need to explain in detail here. Readers can search for information on their own.

It is just a simple schematic diagram above, which leads many junior engineers to take it for granted to use Figure 3 to test the ripple of the power supply. Figure 3 Schematic diagram of incorrect use of the power supply ripple on the oscilloscope side. 51)]The above is probably the method that most junior engineers use when they start testing. This kind of measurement often results in larger power ripple and noise values, which cannot show the real situation as much as possible. Note the wording here. In fact, the oscilloscope cannot fully show the real situation, and there is a certain deviation. As long as it is within the tolerance range, it is acceptable.

Figure 4 shows the consequences of using an oscilloscope with an incorrect measurement method. Since the loop area formed by the ground wire and the probe is too large (the area formed by the cross-section line), it is equivalent to an "antenna" and is easily interfered by EMI. It also absorbs other high-frequency noise in the air, so the output ripple and noise voltage are quite large.

Figure 5 is a schematic diagram of measurement using a dedicated oscilloscope measurement probe. The spring winding on the probe is usually included when you buy an oscilloscope. Take off the ground wire with an alligator clip on the commonly used oscilloscope probe, and then put on the spring winding. This will form a very small loop between the ground wire and the probe, minimizing the risk of noise entering. Figure 23.6 is the oscilloscope probe used by the author. Figure 5 Correct way to use the oscilloscope probe Figure 6 Actual oscilloscope connection method "]After the oscilloscope probe is ready, the next step is to adjust the oscilloscope parameters. If you are measuring ripple, you need to limit the bandwidth, usually to 20MHz. If you are measuring noise, you do not need to limit the bandwidth and need to set the full frequency band.

Set the coupling mode for the selected channel. Because of the ripple measurement, set the AC coupling mode;

Set the bandwidth for the selected channel (assuming ripple measurement), limited to 20MHz. Generally, there will be a bandwidth limit option in the channel settings;

The trigger mode needs to be set to edge trigger mode, rising edge is sufficient, and select automatic trigger to observe the ripple in real time;

Set appropriate sampling lengths for the Y-axis and X-axis. I usually set the Y-axis to 100mV/ and the X-axis to 100us/;

Select peak-to-peak measurement as the measurement method.

(51, 51, 51)]Set the coupling mode for the selected channel. Because of ripple measurement, set AC coupling mode;

Set the bandwidth for the selected channel (assuming ripple measurement), limit it to 20MHz, usually there will be a bandwidth limit option in the channel setting;

The trigger mode needs to be set to edge trigger mode, rising edge is fine, and select automatic trigger to observe ripple in real time;

Set appropriate sampling lengths for the Y and X axes. I usually set the Y axis to 100mV/ and the X axis to 100us/;

Select peak-to-peak measurement as the measurement method.

(51, 51, 51)]Set the coupling mode for the selected channel. Because of ripple measurement, set AC coupling mode;

Set the bandwidth for the selected channel (assuming ripple measurement), limit it to 20MHz, usually there will be a bandwidth limit option in the channel setting;

The trigger mode needs to be set to edge trigger mode, rising edge is fine, and select automatic trigger to observe ripple in real time;

Set appropriate sampling lengths for the Y and X axes. I usually set the Y axis to 100mV/ and the X axis to 100us/;

Select peak-to-peak measurement as the measurement method.