How do parallel AC input resistors reduce conducted EMI?

[BilTy2]

Hello,

Consider the R1/R2 in the attached schematic from Power integrations.. They are described as forming an EMI filter together with C1.

I'd be grateful if someone could explain the mechanism and how effective they likely are, as I'm not very familiar with EMI/EMC. I guess this is a cheaper alternative to the typical low pass Pi filter that I'd expect to see here. Power dissipation would usually make such a setup nonviable, but for this very low power design, they only burn about 7mW at full load @ 85V input.

Aside from the functionality, this doesn't seem like a great arrangement from a safety perspective. RF1 is a 2W flameproof fuseable resistor for protection, but it requires 0.5A just to get to 2W, by which point R1/R2 would have exploded (1/2W vs 55W). Am I missing something?

https://www.power.com/sites/default/files/documents/der-507.pdf

 

[Easy peasy]

Well - they do give 220 ohms of CM, and 440 ohms of DM attenuation, and limit the turn on surge current to C1 to under an amp.

BTW this circuit is illegal in many countries as it may draw more than 5mA ave DC from the mains supply, causing local earth stake corrosion - so read up on standards in the countries of sale ! Better to have a bridge rectifier any how

--- Updated Jul 30, 2024 ---

RF1 is of course dumb in this context and is likely a thoughtless " copy over " from another design

--- Updated Jul 30, 2024 ---

oh yes - D1 can see > 750V pk - another good reason for a bridge.

--- Updated Jul 30, 2024 ---

L1 needs to be rated for > 400v end to end - most cheapies aren't - they go poof after a while . . .

 

[hiphei]

I think r1, r2 and c1 act also as voltage divider for rf frequency signals not to enter the supply from mains direction..though other way around for 'outgoing' emi it is exactly like Easy Peasy wrote.. I think Rf1 should anyway be non flammable type if it is being used in stead of fuse..it's just as dummy way to protect the supply like some old Philips tube tabletop radios have only the transformer overheat protector skipping fuse altogether..

 

[BilTy2]

Thanks for the helpful responses and highlighting the DC issue. It's not something I was aware of.

I'm not using this specific design, but am working on a similar circuit (mains -> 5VDC buck with a half-wave rect.), but avg output is only 50mW, so it would be difficult to hit 5mA avg on the input, even with a terrible power factor, although as you say probably better to have a full rectifier anyway, and likely cheaper given reduced bulk input caps.

My inductor is 500VAC rated.

RF1 is non-flammable fuseable type, but my point was that the current required to break the fuse would have long since turned the 220R resistors into smoke, although I guess it's not impossible that they could fail short.

 

[KlausST]

Hi,

just a remark...

You say "parallel ... resistors".
--> electrically they are not parallel, but in series.

Klaus

 

[BilTy2]

Hi,

just a remark...

You say "parallel ... resistors".
--> electrically they are not parallel, but in series.

Klaus

You are of course correct Klaus. Perhaps symmetric would have been a more appropriate term. I was simply highlighting that equal value resistors are placed on the live and neutral lines, as shown in the schematic.