TPS565201: How to Reduce Input Ripple

Hi,

Please use a 10x probe instead of a 100x probe first. Then you should use small ring instead of GND clip to measure ripple on terminals of C62 or C63. This can help reduce coupling radiation noise. And the layout is not good because input cap Vin side is close to the device but GND side is far from the device, so you need to change the layout. Thanks.

Hi，I use a 1x probe，Testing Coaxial cable was used.

I failed to understand the layout suggestion, can you explain it further please?

Hi,

I mean there's a parasitic inductance from GND pin of device to GND terminal of the input cap. If you use a coaxial cable, why the channel setup is "100x" instead of "1x"?

Hi brother，Is it the green path? Am I understanding it correctly?

100X is ch1，Input ripple is Locked at CH2，CH2 is 1x；

Hi brother, are you a FAE in China?

Hi,

Yes, that's what I mean. So Do you set the bandwidth limit of CH2 to 20MHz? If not, coupling radiation noise will show on the screen. And another way you can try is add a 5Ω resistor in series with Cboot, this can slow down SW rising slew rate.

Can Cboot be reduced？0.1uF to 0.47uF，I've tried adding beads or inductors to the inputs, and that can help as well

I think change Cboot should not be useful. And if adding beads can help, I think that should be conducted interference.

嗨，我在 cboot 中添加了一个 5R 电阻，测试没有改善，

带宽限制为 20Mhz，对尖峰进行滤波。

Hi, 

Could you measure SW waveform and measure rising time before and after adding 5Ω resistor? I think the action should be useful. Please note that the resistor should be in series with Cboot, not in parallel with it. And you can see this document. Thanks.

关于降压转换器中自举电阻器的设计注意事项.pdf

yeah,It doesn't look like much has changed.Thanks！

OK. Do you turn off bandwidth limit when measure this waveform?

close the bandwith limit 20Mhz，Nothing has changed in comparison.

add a 22uh  inductors in input，it's goog，But I don't know why.

AD/DC is supply noise from output，add a 22uH inductors in output，it's very good.why?I  can't explain ti.

I add a 2Ω resistor in output 12V，used to be observe the current，yeah，current wave look like oscillation at DCDC up-MOS open.

Hi,

Sorry I cannot understand what you do in above two reply, could you explain it with schematic? And what issue you want to solve? Thanks.

add a 22uh inductor in DCDC input， The spikes are gone.

But there are other spikes in the input ripple，add a 22uh inductor in AC/DC output，The spikes are gone.DCDC power supply from AC/DC output.