LM5160-Q1: EMC Improvement Review Request

Hi CHO,

Thank you for posting.  Sorry for the delay due to the e2e maintenance.  I will get back to you by COB next Tue USA time.

Best Regards,

Youhao

Hi CHO,

Sorry for the delay.  Can they try to place a 1nF to 2.2nF ceramic capacitor (2kV rating?) across the primary and secondary grounds across the isolation boundary, like shown in our datasheet example schematic (copy and pasted below):

Thanks,

Youhao

Hi Youhao,

The customer had previously applied a capacitor.

However, I deleted it because EMC had a problem.

I tested it by adding 4.7nF according to your advice, but it didn't improve.

It's a system block, and I hope it can be used for your review.

Hi CHO,  thank you and I understand your reason of removing the cross-boundary capacitor because of CMTI.

I see the emission just exceeded the spec at slighter higher than 1MHz, and it should be related to the fast dv/dt on the switching current. Add some 10nF to 100nF at the input filter, and maybe a small ferrite bead in the front, should be able to help.

The block diagram looks like one LM5160 produces all these bias voltages for the three phase isolated drivers. Is that true?  Or each of the isolated driver is supplied by a dedicated LM5160, and there are multiple LM5160 devices used?  I intend to believe it is the latter case according to the schematic shared in your first post. Then, which LM5160 in the system has that EMI emission also shown in you first post? All four LM5160 behave similarly, or just one of them?  Could you single out which one? 

Best Regards,

Youhao