Power supply evaluation practice

Have you ever had the experience of measuring a waveform and seeing that the results changed each time you measured it? How you use the probe when measuring a waveform can make a big difference in the results.

Judging by waveforms measured by bad probing, you run the risk of being forced to install overly costly filter circuits or not realizing that you're out of specification for downstream devices.

This time, I would like to actually use a normal passive probe and a power rail probe for power supply evaluation to evaluate the output ripple level of a step-down power supply.

Evaluation environment

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

Ripple voltage measurement result

This time, we compared the results of measuring the ripple voltage of switching regulators using three types of probes.

Using a power rail probe (Fig. 2)

peak to peak が約7mVと大幅に高周波ノイズの影響をおさえ、より正確なリップル特性を取得できました。

A power rail probe is effective when checking the ripple characteristics of a voltage line that requires a highly accurate low voltage of less than 1.0V.

Use GND spring (Fig. 3)

Better than the GND lead, but still affected by high-frequency noise, causing measurement errors depending on the amplitude.
Compared to using a power rail probe (approximately 7mV), the peak to peak is approximately 14mV, approximately twice as large.

Use GND lead (Fig. 4)

The L component of the GND lead is greatly affected by high-frequency noise, and an appropriate ripple waveform cannot be obtained.
Compared to using a power rail probe (approximately 7mV), the peak to peak is approximately 22mV, approximately three times larger.

Figure 2: Power Rail Probe

Figure 3: General purpose probe (with GND spring)

Figure 4: General purpose probe (using GND lead)

Figure 5: Ripple voltage measurement results

At the end

When measured in the right environment, the LT8609S was found to exhibit very good ripple performance.

If the downstream device does not have stringent ripple requirements, it may be possible to reduce the output capacitors a little more, resulting in a lower cost circuit.

Taking a photo of the probing situation at the time of measurement is a good way to ensure reproducibility when re-measuring.

When I measured it this time, I found that the value of the ripple voltage varied by 2 to 3 times depending on the probing method.

Inquiry

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