EMC filtering worsens current harmonics in SMPS?

[eem2am]

smps emc

hello,

In an offline SMPS, we use an EMC filter to reduce conducted emissions.

Is it true that such EMC filtration (eg caps and common mode & differential mode choke) can actually worsen the current harmonics?

 

[dipnirvana]

leakage current in smps

i don't think so , as in EMC in the end you measure current only

 

[eem2am]

harmonic current smps

Thankyou dipnirvana.......you say the same as i think.......its just that i was just telephone interviewed by a Senoir Power Electronics Engineer, and he asked me how i would lessen conducted emissions.....

-i said with an EMC filter.......

-he then said that putting in an EMC filter could reduce the conducted emissions, but also the EMC filter could make some of the current harmonics worse......

....and he was asking me how i could do the EMC filtering, without making the current harmonic situation worse.

I was totally confused as reducing conducted emissions should reduce the current harmonics ?

Maybe he was thinking about the accidental excitement of resonant effects in EMC filters ?

 

[dipnirvana]

current harmonics

Well resonance will happen only on selected frequency. May be they were talking of harmonics of base AC signal (50Hz) as PF will change from "1"

 

[FvM]

emc harmonics

but also the EMC filter could make some of the current harmonics worse....

This sounds somewhat confusing to my opinion. As dipnirvana mentioned, EMC filters can slightly decrease the power factor by adding capacitive apparent power. But it don't generate harmonics (Not considering bad designed filters with saturated inductors...). The question is, if in special cases, it may increase the harmonic currents of a nonlinear load, e.g. a non-PFC power supply? I wouldn't expect that generally.

As another point, the harmonic currents measured at an EMC filter input can be higher than at the output, if considerable voltage harmonics are present. But that can't be hold against the instrument utilizing the filter and isn't a problem in EMC measurements

 

[eem2am]

emc filter mains

EMC filters can slightly decrease the power factor by adding capacitive apparent power

...i am very doubtful that an SMPS's EMC filter would add capacitive apparent power.

this is because leakage current limitations always mean that the capacitors used in EMC filters are pretty low value...and the main means of filtering is via the differential mode inductance......so the EMC filter looks mostly inductive.

But the thing is that is killing me is that , an active power factor corrector will actually mean that the current drawn FROM the EMC FILTER is in phase with the voltage at the output of the EMC FILTER.

This is, as you know, NOT the required result of the mains input current being in phase with the mains input voltage..................

so yes you are right in that you say that the nasty EMC filter is going to screw up the work of the active power factor corrector....but it will make the SMPS look inductive..........and if its a decent level of filtration, it will look very inductive..........which is bad news.

Again though, this is not creating harmonics, but rather phase shift, so i still dont know what the engineer who i spoke to was talking about

 

[FvM]

reducing conducted emissions

this is because leakage current limitations always mean that the capacitors used in EMC filters are pretty low value

Not necesarily. If you think of a differential mode filter, involving a "X"-capacitor, we are talking of .1 upto 10 uF in case of a high power SMPS.

the main means of filtering is via the differential mode inductance......so the EMC filter looks mostly inductive.

They "look"inductive at the switching frequency, but most likely capacitive at the mains frequency and lower harmonics.

But the thing is that is killing me is that , an active power factor corrector will actually mean that the current drawn FROM the EMC FILTER is in phase with the voltage at the output of the EMC FILTER.

To my opinion, the difference between filter input and output power factor is mainly caused by the capacitive current rather than the inductive voltage drop. Depending on the PFC design, you can compensate for the capacitive current, achieving ideally unity power factor at the filter input.

so i still dont know what the engineer who i spoke to was talking about

This is unclear to me as well.

 

[eem2am]

smps main caps

actually its all becoming clear i think?

Some people snuff out EMC conducted noise by using huge differential mode chokes...these have interwinding capacitance problems which lets the noise throuigh tto the mains instead of blocking it away