MPM3695 input noise

Hello,

I experience a lot of noise at the input of the module. This parasite injected on the 12V input rail is critical and disturbs the other components on the board. The parasite appears at the switching frequency of the converter, as illustrated in the following figure (green).

Here is a zoom of the noise, measured with a differential probe at the input pin of the converter (the parasite is not present at the ouput) (blue).

I have isolated the DCDC from other DCDCs on the board to confirm that the noise is injected by the MPM3695. The decoupling at the input consists of 3 ceramic capacitors (22µF + 1µF), I tried to change the input caps (adding nF and pF high frequency caps, electroytic with larger ESR,…) but nothing really had significant impact. The design is close to what is recommended in the datasheet, that is why I have some difficulties to understand what happen.

Did anyone experienced the same phenomenon?

The schematic implementation of the converter is depicted below. (3 image in one because I’m restricted to upload only one image)

All switching power devices crest high speed dledge noise. This is similar to the output side noise induced by the ASIC and the PDN impedance.

I don’t know what your board design looks like, but you’ll want to do EM simulation of the board and converter and probably design an appropriated filter.

If you need help, email me at steve@picotest.com

try putting a ferrite bead or even an inductor in series with the input power to this device. Not clear, I mean leave the caps as they are but add a series L between the schematic shown and the system power