TPS63802: Output noise issue

Hello.

The measurement was not done by me.

Maybe this noise is due to a bad measurement method and also maybe due a not very good oscilloscope. Like due to the negative clip of the probe, which is not very short.

Hello.

On the output there is a sequence of 3 capacitors.

Now the sequence is 47uF - 47uF - 47uF

We will also test soon, with these sequences:

47uF - 100nF - 47uF

47uF - 100nF - 1uF

Hi.

For the 3 output capacitors, we did the following sequence: 47uF, 100nF, 1uF. But there is still a lot of noise. 

Regards.

Hello Jeferson,

As noise can be picked up by the scope probe, a picture of the setup would be interesting.

Have a look at this app note for more details: How to measure ripple for better design outcomes (ti.com)

Best regards,
Brigitte

Thank you.

We managed to reduce the noises.

In the original PCB, at the output of TPS63802 there is 3 capacitors: 47uF, 47uF, 47uF, this sequence. All the 3 capacitors are the same PN. According to the webpage of this capacitor's PN, the resonant frequency of it is around 800-900KHz. And it was observed, measuring later with the correct setup, using FFT, a peak of noise at around 1.7~1.8MHz, shown below, which is a frequency close to the switching frequency of this converter.

Then we changed the sequence of output capacitors to 22uF, 22uF and 47uF, because the resonant frequency of the 22uF is closer (than the 47uF one) to the peak frequency of the noise:

>>> 47uF / 10V / 0805

>>> 22uF / 16V / 0805 (resonant frequency closer to the noise frequency and closer to the switching frequency of the converter)

And at the end of all tests, we managed to obtain much better result, much lower noise at the output.

Thank you.

Hi  Jeferson,

Congratulations. 

Should I close this thread since seems no further support needed? 

Reagrds

Tao

Hello.

Yes, can close.

Regards.