LM25005: EMC problems with design from webbench

Andreas Wileur

Part Number: LM25005

Hi,

I have a 3.3V regulator design with LM25005 straight from Webbench. Power wise it delivers exactly as expected, ripple is low everything works fine. But the other day we did our preliminary EMC-tests and found a lot of emissions that we eventually traced to the LM25005, as everything else on the board is in reset or disconnected. The emissions are quite broad band and get worse the more we load it. In the attached screenshot the load is ca 1A through a 3.3Ohm resistor.

And this is without any load:

This is the schematic:

We have tried several things with various results:

1) Removed C121 made no difference

2) Putting a snubber network across D14 made some difference, but we had to use ca 4Ohms and 10n to get acceptable levels, which is a lot more than the datasheet recommends.

3) Changing the diode D14 also made a difference, but mostly at what frequency the emissions occur.

4) Increasing the inductance from 10uH to ca 100uH made no measurable difference.

Any help is very welcome!

Best regards,
Andreas Wileur

over 5 years ago

0 Katelyn Wiggenhorn over 5 years ago

TI__Genius 11040 points

Hello Andreas,

Your images did not attach to the previous post. Would you be able to reshare the images?

Would you also be able to share your layout - as the layout is very critical for EMI performance?

Best Regards,
Katelyn Wiggenhorn

0 Katelyn Wiggenhorn over 5 years ago in reply to Katelyn Wiggenhorn

TI__Genius 11040 points

Hello,

I want to follow up on this post. Would you be able to reattach the images and share the layout as well?

Thank you,
Katelyn

0 Andreas Wileur over 5 years ago in reply to Katelyn Wiggenhorn

Prodigy 60 points

Hi, I still have problems attaching the pictures, but I think the diode was the culprit. I picked one that probably was too fast and after a lot of testing other options I have found one that produces generates less EMI. Another round in the EMC-lab will tell whether it's really solved.

Webbench produces nice simulations and the voltage/current ripple was spot on, but it doesn't take all parameters into account. Also, it's hard to compare diode datasheets from different manufacturers since they highlight different parameters.

I originally used SBRT3U60SA, chosen for it's low voltage drop and therefore high efficiency. I finally settled on B370-13-F, possibly with a small snubber circuit. We'll see next week.

Thanks for following up!

Best regards,
Andreas Wileur

0 Katelyn Wiggenhorn over 5 years ago in reply to Andreas Wileur

TI__Genius 11040 points

Hello,

I am happy to hear that the problem is resolved. I am going to close this thread, but please feel free to open another thread if you have any additional questions.

Best Regards,
Katelyn

Power management

Power management forum

LM25005: EMC problems with design from webbench