TPS53318: the noise of TPS53318 when FCC/SRRC RE certification

Wade shao

Part Number: TPS53318

Background

The TPS53318 in Customer’s Main-board, and the TPS53318 output a power supply (4.2V) to Main-borad.

In the RE certification process of FCC/SRRC, at 100MHz-500MHz, some noise be detected in main-board, and so, the certification failed.

Debug

1, Noise can be detected in all the parts of main-board. And on the right above of TPS53318, The strongest noise be detected, the difference is about 20dB. So customer suspect TPS53318 is the source of noise.

2, Replace TPS53318 by DC-power directly, the noise disappeared.

3, the customer tested FCCM and Skip mode separately, The noise situation is the same.

Need help

1, Why does noise occur?

2, What can we do to make the noise go away?

over 6 years ago

0 JohnTucker over 6 years ago

TI__Guru**** 191445 points

Wade,

Can you share details? TPS53318 is a switch mode dc/dc converter. Any switch mode converter may generate high frequency components. Can you post the schematic, PCB layout and a waveform showing the switching node voltage? Make sure to zoom in such that any overshoot and ringing can be observed (two switching cycles should be sufficient).

0 JohnTucker over 6 years ago in reply to JohnTucker

TI__Guru**** 191445 points

Wade,
Any updates?

0 Wade shao over 6 years ago in reply to JohnTucker

Prodigy 180 points

Dear,

The customer has shared the schematic, But PCB layout and waveform are temporarily unavailable. I will share it as soon as possible.

In the schematic, V3S = 9V ~ 13.2V .

thanks.

wade

0 JohnTucker over 6 years ago in reply to Wade shao

TI__Guru**** 191445 points

Wade,

While we wait, did it fail radiated or conducted emissions?

0 Wade shao over 6 years ago in reply to JohnTucker

Prodigy 180 points

0 JohnTucker over 6 years ago in reply to Wade shao

TI__Guru**** 191445 points

Wade,

You state the design fails between 100 MHz and 500 MHz. You will need to zoom in on various waveforms like Vin and SW an look for switching artifacts in that frequency range. The will usually occur at the rising and falling edges. The waveforms above show DCM operation. the resonant tank of the output filter is ringing at about 8 MHz. I doubt that is the source of the issue. I do not see any snubber on SW or series resistor on VBST. You may need additional filtering on Vout and or Vin if you are failing conducted emissions.

0 JohnTucker over 6 years ago in reply to JohnTucker

TI__Guru**** 191445 points

Wade,

I will temporarily close this action. If you get more details and need further assistance, please feel free to re-open this by responding.

0 Wade shao over 6 years ago in reply to JohnTucker

Prodigy 180 points

Dears,

thanks a lot

the customer has added RC between LL and Ground, and then, the noise about at 200MHz has been reduced.

but the noise at 300MHz is still exist.

please refer to the first and the second pictures.

And, I will advise the customer to add Register at VBST,

thanks again.

0 JohnTucker over 6 years ago in reply to Wade shao

TI__Guru**** 191445 points

Wade,

Did you check the time domain waveforms? How were the values for the snubber determined? It may not be optimized. Also layout is important for snubber placement.