LM25116: PCB review

Yue TANG

Part Number: LM25116
Other Parts Discussed in Thread: CSD18537NQ5A, CSD18563Q5A

Hi Champs,

My customer use LM25116 in their new design. but recently, they met a EMI issue which will disturb their USB port. need your help to check their PCB design to see if something wrong. thanks.

Their design target is 24V to 5V 9A,. Schematic and PCB is as attached.

SKYLAKE _ 61-24V-5V.pdf

PCB_SASBY_ISLAND_VER_0924p2.rar

over 6 years ago

0 Timothy Hegarty over 6 years ago

TI__Guru* 87737 points

Hi Yue,

Please send a completed quickstart file (download it from the LM25116 product folder) for this design so we can verify component selection. Have you taken an EMI sweep to determine what frequencies might be problematic?

See part 6 of this article for more info in layout an EMI mitigation for controllers:

http://www.how2power.com/other/EMI_Guide.php

0 Yue TANG over 6 years ago in reply to Timothy Hegarty

TI__Genius 9045 points

Timothy,

What is quickstart file? is it a kind of checklist? I can't find this file from your link. could you please provide a more specific link?

please just help to check the layout to identify the potential issue? their schematic based on WEBENCH, and have reviewed in another E2E thread. should not have big problem.

0 Timothy Hegarty over 6 years ago in reply to Yue TANG

TI__Guru* 87737 points

http://www.ti.com/tool/LM5116-5116WG-25116DESIGN-CALC

0 Yue TANG over 6 years ago in reply to Timothy Hegarty

TI__Genius 9045 points

Timothy,

Customer updated the xls file to me. please help to check . thanks.

8737.LM5116_quickstart.xls

0 Timothy Hegarty over 6 years ago in reply to Yue TANG

TI__Guru* 87737 points

Hi Yue,

Changing the ripple current % in the quickstart file gives an inductor value of 3.7uH that matches the schematic. The recommended Cramp then becomes 255pF (choose 270pF standard value, currently 330pF in the schematic).

Also note that the Miller plateau of the high-side FET is ~4V (see Fig. 4 of the CSD18537NQ5A MOSFET datasheet). This is a bit high if the gate drive is only 5V (as Vout is connected to VCCX of the LM25116). Please ensure there is adequate gate voltage overdrive for the FETs. Note that the low-side MOSFET (CSD18563Q5A) has an Rdson rated for Vgs = 4.5V, but the CSD18537NQ5A is rated only with Vgs = 6V. One option here is to use the CSD18563Q5A for both FETs.

In terms of EMI, I notice there is no common-mode input filtering (e.g. common-mode choke or ferrite bead in the two input lines). This is typically required to mitigate high-frequency emissions greater than 10-20MHz.

Regards,

Tim

0 Yue TANG over 6 years ago in reply to Timothy Hegarty

TI__Genius 9045 points

Tim,

Thanks for the schematic check, could you please kindly check the PCB? thanks.

0 Timothy Hegarty over 6 years ago in reply to Yue TANG

TI__Guru* 87737 points

Hi Yue,

This is a very large board layout. We will need to setup a call to review if needed. I recommend installing a common-mode choke at the input and see if that fixes the issue.

Regards,

Tim

0 Yue TANG over 6 years ago in reply to Timothy Hegarty

TI__Genius 9045 points

Tim,

I'll send you a mail to further discussion. thanks

Power management

Power management forum

LM25116: PCB review