It is a well known fact that offline SMPS in a non-metallic enclosure will fail Radiated Emissions.
Usually, plastic enclosed offline SMPS have a metal foil lining for shielding to allow it to pass Radiated Emissions.
No one thinks its strange that we shield things like comms cables...the di/dt, and the dv/dt can be far worse in an offline SMPS than in low voltage comms...so it obviously needs shielding aswell.
I went to one co who did not shield a 70W offline flyback....they failed radiated EMC.
Best way to pass radiated EMC for offline SMPS....is to shield the SMPS in earthed metal or metal foil (can have holes and gaps).....the earth should come from the mains cable.
Many products of course, cheat......and dont bother certifying for Radiated EMC.
Once we drilled a hole in a metal box which contained an open frame offline SMPS....we brought a cable out from the offline SMPS, and took it to a charger box outside of the metal box......and we damned well needed to shield that cable......Radiated EMC was fail fail fail when we didnt shield that cable!
Why is it that people do not appreciate that an offline SMPS will fail Radiated EMC if it is not shielded in metal?...an offline SMPS, with its high voltages, and hard switching PFC stage, is the most violent thing known in the world of EMC.........and people dont think it needs shielding, when other far less "EMC violent" things do need shielding.
Is anyone going to pretend that a low voltage comms cable, or a 3V3 micro running at a few MHz, is going to be more "EMC offensive" than a hard switching offline switcher?
Next time one passes a plastic fire panel.........ring up the manufacturer...they will tell you, when those fire panels got converted to plastic case, instead of metal case...the open frame offline SMPS's inside tham had to get put in metal enclosures.........why does the industry try so hard to cloak this in secrecy?........is it because mitigating hardware for passing radiated EMC is more expensive?.....is it because getting through Radiated compliance testing is extremely expensive?
Or is it because nobody fines you if you do produce something that fails radiated EMC.....after all, imagine how long it would take the authorities to do a radiated scan test on every product on the market.
Your competitor may threaten to report you if they see that you fail radiated on the competing prodct....but then they dont,, becauase everyone just declares "amnesty", and promises not to report competitors, as long as they dont report them for their own Radiated EMC fails.