TPS40170: Noise Issue on My power supply

You should use the short circuit to do the test, and set the bandwidth in 20MHz. the long GND cable of the probe will bring some noise because of the antenna effect

It is not clear to me what the purple trace is.  I am guessing you have "persistence" turned on.  I am also guessing that you are complaining about switching noise.  You have to be careful to measure this correctly.  You need to measure directly across your output capacitor with a very short (< 1 cm) ground lead, or better still use the "tip and barrel" method.  If the noise is still too large, there are a few things you can do. 

First off you should consider some high frequency bypass capacitors in parallel with your main output filter capacitor, 0.1 uF, 0.01 uF, etc.

Check your PCB layout to make sure there are no unnecessarily long circulating loops.

You can consider an RC snubber from the SW node to GND.

Let us know if this helps.

!5 V p-p seems to be not reasonable.  Are you sure you are measuring it correctly?  Did you include high frequency bypass capacitors and or RC snubber?

I changed the output capacitor from 100uF to 5 x 22uF. Can you please check the attached wave forms.

I checked the snubber circuit designing theory, it is saying need to know the ringing frequency. How can i find the ringing frequency ? Or any other specified values can used for this snubber ?

D you use the right way to test the ripple?

For the snubber circuit, you can catch the waveform of SW to see the frequency of ring, and then add a cap in parallel with low side-MOS to make the frequency of ring to half , then you can know the parasitical inductance, finally you can get the value of the resistor.

Enclosed is the procedure that I use for RC snubber design,

8836.RC Snubber.pdf

I added snubber circuit parallel to the out put mosfet and captured the waveforms. But still  noise increasing while load increases.

5v output waveforms after snubber added.zip

Hello Sir,

I added a 15R resistor at the gate of two mosfet's now the noise is reduced, is it a right way ?. Plaese check the attached waveform having load(1.5A+Raspberry pi).

Hello,

We made a two layer board, here i am attaching the gerbers, can you please check it ?

I changed the resistance value from 15R to 10R, then the noise increased to 240mV(Switching peak) at about 2A load.

How i can reduce it further, do i need to change the mosfet ?

Carrier_FAB_REVA_12APR2017.zip

Sorry, we don't have these software.

For the mos, you can refer to User's Guide. But I still think you had better change your layout. In fact, the improved by MOS is extremely limited. I think your main problem is the layout of two layers.