2
\$\begingroup\$

The 30 V 500 mA power supply I'm working on needs to pass the EMC (EN55015) test. I am attaching the necessary circuit diagram (common mode and differential mode filters) to pass the test according to my research and trial results.

It passes according to the results of the tests we have done in our own laboratory environment, but I also want to suppress 10 dBuV between 0.7 and 1.8 MHz. (Test result and EMC filter are in the same photo.)

I did a lot of research and experimentation but couldn't find a clear result. I need your advice on this.

enter image description here

\$\endgroup\$
5
  • 1
    \$\begingroup\$ ~1 MHz is fairly straight forward to simulate if you add the ESR and ESL of your capacitors, equivalent parallell capacitance (EPC?) and ESR of your inductors, leakage inductance of magnetics, stray capacitance of your tracks and simplify the entire DC/DC stage to just pull the same dI/dt in one cycle, you should be able to capture the culprit. \$\endgroup\$
    – winny
    Jun 27, 2022 at 13:29
  • \$\begingroup\$ Is this a conducted emissions (CE) requirement/test? \$\endgroup\$
    – SteveSh
    Jun 27, 2022 at 14:04
  • \$\begingroup\$ Do you have PE (earth) connection? Where's this measurement taken from? L only or N only? Have you tested for both terminals? Also do you have any info about if this is a CM or DM noise? \$\endgroup\$ Jun 27, 2022 at 14:07
  • \$\begingroup\$ This is measurement taken from N. But I measured both L and N. The report is the same on both.From this result I understand that the noise source is common mode. Is this true?@RohatKılıç \$\endgroup\$ Jun 27, 2022 at 14:59
  • \$\begingroup\$ Yes, it is. @SteveSh \$\endgroup\$ Jun 27, 2022 at 15:00

1 Answer 1

1
\$\begingroup\$

Even the rectifier may be a problem.

Use ultra-fast diodes for the diode bridge. Regular "good old rectifiers" are a pit of despair when it comes to EMI. It might sound counter-intuitive: how are "fast" diodes less noisy than good old "slow" diodes? The "slow" diodes still turn off relatively quickly, they just do it too late: they should be switching close to zero voltage to be quiet, and the later they switch, the more voltage there's across the switch - and also the more current. You'll see an improvement just from better diodes - perhaps not in the frequencies you are concerned about, but it will make life better.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.