Ripple waveform measurement

In the previous technical article, "Is that measurement correct? The true ripple voltage was 1/3", I explained that the ripple voltage value differs depending on the probe.

As for the ripple measurement result at this time, we were able to observe a beautiful ripple like the waveform in Fig. 1. This proves that power rail probe measurements are valid.

However, until I got this result, I was in trouble because I was suffering from noise and could not acquire a clean waveform.

Figure 1: Ripple voltage measurement results

Initial measurement environment and ripple waveform

This time, we used the LT8609S evaluation board, which is a 42V withstand voltage synchronous rectifier MOSFET built-in step-down DCDC from Analog Devices.

Evaluation board settings

The change from the original evaluation board is to change the feedback resistor R5 from 182KΩ to 300KΩ and set the output voltage to 3.3V.

Evaluation conditions: Input voltage 12V, output voltage 3.3V, load current 2A, switching frequency 2MHz setting

Figure 2: LT8609S Evaluation Board Schematic

Ripple voltage measurement result

This time, I used two types of probes to measure the ripple voltage of a switching regulator.

A waveform like the one in Fig. 3 was obtained. Due to the noise that appears in the red frame, even if you try to obtain peak-to-peak measurement results, you will not be able to measure the correct value.

In the measurement using the GND lead, the GND loop becomes large and the effect of noise is large. Even power rail probes, which are less susceptible to noise, are affected by noise.

I reviewed the probing method many times, but it didn't improve, so I finally changed the power supply to the battery prepared for in-vehicle use and experimented.

Figure 3: Ripple measurement results

Figure 4: Power Rail Probe

Figure 5: General purpose probe (using GND lead)

Replace the stabilized power supply with a 12V battery and measure

I changed the power supply from commercial power supply → stabilized power supply → evaluation board to battery → evaluation board connection.

 

Figure 6 shows ripple waveform results measured with a power rail probe when powered by a battery.

 

The noise that was in the shape until a while ago disappeared cleanly.

 

Figure 6: Ripple waveform when powered by battery

Summary

The ripple noise source was found to be noise from the mains or regulated power supply.

(Unfortunately, we have not been able to follow the separation of commercial power supply and stabilized power supply.)

In the future, there will be more opportunities to measure low-voltage ripple, so we were able to reconfirm the need to pay attention to input power supply noise when performing evaluations.

This time, there was a battery at the measurement site, so we were able to isolate the noise source.

As in this example, noise may be picked up by the measurement environment other than the measurement equipment, and the measurement may not be performed correctly, so caution is required.

In addition, we often receive requests from customers to do something about the noise in their switching regulators, as power supply noise degrades the signal quality of FPGAs. It was a good experience for me to say that I was troubled by power supply noise in the measurement environment, and I would like to propose and support future customer support so that they can design power supplies with less noise.

Supplement: In the experiment in the technical article "Is that measurement correct? The true ripple voltage was 1/3", I gave up the stabilized power supply and measured it with a battery.

Inquiry

If you have any questions regarding this product, please contact us using the form below.