SMPS EMC filter on AC or DC side?

Hello,

In an offline SMPS , would it not be better to put ALL of the EMC filtration “after” the Mains rectifier (DC side) instead of putting some on the AC side?

I mean, the common mode choke and the X caps should best go on the DC side of the filter?

This SMPS schematic has a common mode choke and a 330nF EMC capacitor on the AC side, which would appear to be a complete waste of time. ?

Offline SMPS schematic
**************


I can see that an amount of very high frequency filtration (such as ferrite cylinders or beads) around the AC input cable might be needed,
-but certainly not the ~100nF X capacitors and ~10mH common mode chokes which are commonly seen on the AC side of most offline SMPS’s.

After all, we want the high frequency switching harmonics to be traversing as small a loop as possible, so what’s the point in putting the mains rectifier in that loop and making it bigger than it needs to be?

The ONLY reason that I can think of for putting a filter on the AC side of the rectifier is to dampen down the peak of any overvoltage transient such that the rectifier and other components are slightly protected from it.

However, unless an AC-side filter was unreasonably large, it would not really give enough attenuation of an overvoltage peak anyway. ?

So, taking that into account, the ONLY reason that I can now think of to have an AC side filter is to slow up an overvoltage transient’s rising edge prior to clamping it with a MOV or transorb (the MOV or transorb also being placed on the AC side).

-The AC-side filter would be useful in this case because MOV’s and transorbs are relatively slow to turn ON, and so would benefit from having some filtration before them.

Otherwise, I can see utterly NO reason to have an AC side filter in an SMPS.

Both X capacitors AND common-mode-choke would surely be better situated on the DC-side of the mains rectifier ?

i believe this is to prevent transients from entering the indoor power net and causing problems there.

thankyou steinar96,

though please could you explain, i would be very grateful, as i don't see how the SMPS can cause a transient which could effect the indoor power net.

also, even if the SMPS does cause some transient event, the filter would still be better placed on the DC side of the mains rectifier?

i think filter on main is much needed for EMC, as for a example, if you use that power supply and also a radio/connected on same socket, there would be dome distortion(considering the noise frequency). and in that case your power supply generated a noise,

thankyou leoren_tm


the thing is, it is better to filter the noise away on the DC side of the rectifier -closer to the switch loop where the noise is being caused?

rectifier also a source of noise as switching speed of your PSU,

thankyou leoren_tm,

i expect that the mains rectifier is only a small source of noise....just a small "blip" at 100Hz.

-this would need just a small 1nF X cap on the AC side, and no more. ?

blip? or not
anyway, if you have some EMC test instrument, then that would answer your question, on the amplitude and frequency of noise, if you dont have then maybe by trial and error<--test,

i expect that the mains rectifier is only a small source of noise....just a small "blip" at 100Hz.

Click to expand...

It's a non-PFC rectifier, that generates considerable odd harmonics at 150, 250, 350, ... Hz. But of course, a common mode choke won't have much effect on current harmonics.

would it not be better to put ALL of the EMC filtration “after” the Mains rectifier

Click to expand...

Which advantage do you expect?

first hit on google. **broken link removed**
thanks for info FvM

Quote:
would it not be better to put ALL of the EMC filtration “after” the Mains rectifier

-Which advantage do you expect?
Un-Quote

This way i will confine the SMPS switching frequency harmonics to as small a loop as possible.