LM25118-Q1: EMC RE test fail when LM25118 is enabled

LIN SUN

Part Number: LM25118-Q1

Hi expert:

There's one dcdc lm25118-q1 on my board, this dcdc is designed according to SPEC datasheet.

When doing EMC Radiation Emission (RE) accordingt to CHAPTER6.4 of CHINA GBT18655-2020, board noise cannot be passed once dcdc is enabled.

Could you share some sugestions please? Thanks!

over 4 years ago

0 Brigitte over 4 years ago

TI__Guru 65200 points

Hello Lin,

To get more insights in the issue you are observing, it would be interest to understand at which frequencies you observe issues.

But let me give you some basic information which might help:

1. Often either the input or the output lines are the antennas causing radiation. So increasing the input cap or output cap network. It is important that you add capacitors that are still capacitive at the frequency you want to filter. If these additional filter capacitors can be placed as close as possible to the power stage the antenna area can be reduced.

2. Sometimes the radiation is caused by the inductor, so a change of the inductor can help if a different physical structure is used for the new inductor.

3. If you have series resistors on the input and output side, sometimes exchanging these with ferrite beads can be helpful.

4. If you have gate resistors, changing these might help or if you have snubbers on the resonance diodes, these might help as well if calculated for the problematic frequency.

0 LIN SUN over 4 years ago in reply to Brigitte

Intellectual 620 points

0 Brigitte over 4 years ago in reply to LIN SUN

TI__Guru 65200 points

Hello LIN,

Did my recommendations help solving the issue?

0 LIN SUN over 4 years ago in reply to Brigitte

Intellectual 620 points

RE fail.ppt

0 Brigitte over 4 years ago in reply to LIN SUN

TI__Guru 65200 points

Hello LIN,

In general every switch mode power supply creates high frequency noise. This is the nature of an SMPS. Passing of failing an EMC test is highly dependent on the used external components and the layout.

In addition I expect that there is a load connected to the converter, so if you disable the converter by removing the resistor, you do not supply the load and therefore more than just the converter is disabled, so a simple conclusion like you have in the ppt is not the case, I think.

When I reviewed your schematic it seems that you are using very little input capacitance. The converter together with the load might cause ripple on the input lines which act as antennas.

0 LIN SUN over 4 years ago in reply to Brigitte

Intellectual 620 points

Hi Brigitte

My design is powered by vbat_filter, and vbat_filter has high value capacitors. So Vin should be fine. We are still investigating and thanks.5417.RE fail.ppt

0 LIN SUN over 4 years ago in reply to Brigitte

Intellectual 620 points

Hi Brigitte

My design is powered by vbat_filter, and vbat_filter has high value capacitors. So Vin should be fine. We are still investigating and thanks.0336.RE fail.ppt

0 LIN SUN over 4 years ago in reply to Brigitte

Intellectual 620 points

Hi Brigitte

My design is powered by vbat_filter, and vbat_filter has high value capacitors. So Vin should be fine. We are still investigating and thanks.5466.RE fail.ppt

0 LIN SUN over 4 years ago in reply to Brigitte

Intellectual 620 points

Hi Brigitte

My design is powered by vbat_filter, and vbat_filter has high value capacitors. So Vin should be fine. We are still investigating and thanks.4428.RE fail.ppt

0 Brigitte over 4 years ago in reply to LIN SUN

TI__Guru 65200 points

Hello LIN,

Did any of the recommendations in the first answer help?

0 Brigitte over 4 years ago in reply to Brigitte

TI__Guru 65200 points

Hello LIN,

I expect that one of the recommendations in my first answer solved the issue. If this is not the case, please add another post below or if this thread is already locked, start a new thread.

Power management

Power management forum

LM25118-Q1: EMC RE test fail when LM25118 is enabled