Flyback SMPS: RCD clamp does not solve EMI problem?

[treez]

Hi,
Page 6 of the following, with reference to fig 4b, states that the RCD clamp across the flyback SMPS primary has not solved the EMI problem because the leakage inductance continues to ring in the interval of reverse recovery of the RCD clamp diode.

..This may be true, but during this particular interval, the leakage inductance is ringing with the Capacitor of the RCD clamp. As such, the ringing frequency is reduced to a couple of MHz at most. This low frequency is not likely to cause a radiated emissions problem. As such, surely the RCD clamp has solved the EMI problem?

https://www.ridleyengineering.com/images/phocadownload/12_ flyback_snubber_design.pdf

 

[dick_freebird]

You say "The" EMI problem as if there can
be only one....

 

[Easy peasy]

most major emc issues in the 90 - 300MHz band are almost always caused by the o/p diode - get a really good schotty type and use that - with a snubber ..., and/or slow the turn on of the fet ...

--- Updated Feb 21, 2021 ---

in the RCD snubber on the pri side - if the D part is not a V fast diode - then you can get VHF ringing from that too - if the fet turns on fast

Power integrations has all their chips with a "progressive" gate turn on to try and minimise this effect - they used to mention it in their literature - but have deleted it over time to keep the opposition guessing as to how PI's EMC is better ...

 

[treez]

and/or slow the turn on of the fet ...

Thanks, your comments are great advice and i will heed this whilst i am in the rad emissions lab.
We are using a ICE3A1565 and we unfortunately have no access to the internal FET. The drain node is also the "heatsink node".....gets better

https://www.infineon.com/dgdl/Infineon-COOLSET_F3-DS-v02_03-en.pdf?fileId=db3a304412b407950112b428f6ba3f30

 

[treez]

You say "The" EMI problem as if there can
be only one....

With a flyback, in an earthed enclosure but with exiting cables, the very worse one, i believe you would agree. is High frequency radiation exiting on the cables. Its the hardest to solve, as it can jump round filters etc....and creep through inductor interwinding capacitance.