The power supply is 12V to -5V, BUCK-BOOST topology, high-frequency noise at the test output, and a solution.
Using a 500MHz oscilloscope and a 1G probe, the test has used a ground spring, the yellow line is the noise of the output capacitor to the ground, and the blue line is the SW switching waveform.
There is 400mVpp, 400MHz high frequency noise at the output, how to solve it. The schematic diagram and test waveform are shown in the figure.
Franco Veröffentlicht am January 5, 2021
The source of this noise should be the distributed capacitance of the inductor, which generates a high-frequency pulse current at the leading edge of the switch ON.
Countermeasures:
1 Choose an inductor with a small distributed capacitance
2 Output string beads
Marcellus Veröffentlicht am January 5, 2021
Ripple has been done very well, 400MHz spike interference, you can understand that it is a high-frequency emission, the possibility of spatial coupling is the greatest, the output may be added with a magnetic ring or magnetic beads, and electrical shielding may have to be included to get clean The output waveform.
However, the general digital and analog circuits are completely insensitive to this, and it is not a big problem to use them directly.
Carver Veröffentlicht am January 5, 2021
In fact, most applications are insensitive to noise at this frequency and can be ignored.
