EMI and harmonic filter

For shame, Graham! You didn't answer the final question:

Because differential is not the same as common-mode.

Mark

I think they deal with different frequency ranges. EMI involves radio wave frequencies (say, 30 kHz and above) while harmonic filters deal with the lower multiples of the power line frequency that occur when nonlinear loads like rectifiers are used on sine wave sources. The first kind are there to reduce radiated or conducted interference, while the second are there to reduce the distortion of the power line waveform.

In the same way as differential is not the same as common-mode.

In the same way as differential is not the same as common-mode.

Graham

Whoops ! My apologies. You're quite right of course. ;-)

Graham

hi John, Suppose we are talking about the conducted EMI in the range of well known range 150KHz -30Mhz, how does the differential mode noise differ from harmonics. Is not the harmonics cause of the differential mode noise. Since the Differential mode filter aim is also remove high frequency component from the supply, is not it same as reducing distortion. This thing is really bugging me. I have read several books on EMI, and none explains about it. I wish to talk to some of the authors, unfurtunately easy communication to authors are not mentioned in books. If they are fundamentally different, whats that? if not why so huge literature on such differential filter design. and i do not see any difference from classical filter design. kristo

hi, this discussion was not to annoy anyone. It was merely for learning purpose. The person who himslef do not know the dobut runs to beat the other back. It has been clearly seen from the discussion. The harmonics i was refering is to switching frequency harmonics is SMPS. anyway, A lot of thanks to John for kindly putting his concepts.

If they occur at harmonic multiples of the power line frequency, then they are also harmonics. If they are caused by other sources of energy (switching regulator pulses, for example) then they are not harmonics. But most harmonic filters deal with the lower multiples, because these are what propagate efficiently down the power line and affect other loads.

Yes, they are one kind of differential mode noise. Any source that puts a different voltage on the 2 power lines connecting a piece of equipment to the power lines is a source of differential mode noise, but not all are harmonically related to the power line frequency.

Sure. At the lower harmonics of the power line frequency, you don't want to cause such distortion, because it affects other loads on the same line. If you have a sensitive load, you may need to remove some of this distortion to protect your load. The higher end of the frequency range (where it is not so important or obvious whether a frequency is an exact multiple of the line frequency or not) involves preventing radio frequency radiation leaking out of a piece of equipment (where the power line acts as a transmission line) or problems with such energy getting into your load (but usually the first). Differential and harmonic filters do these things. Most differential mode filters are either there as part of a spike suppression system (for line spikes coming in when big load switch off or for distant lightning disturbances) or are part of a conducted noise suppression system (to meet standards).

The big problem with differential mode filters that must be solved is that they have to pass a really big differential signal -- the line voltage and load current. Common mode filters can often be much more compact, because they can use two winding coupled inductors that cancel the magnetic field of the load current (that big, required differential signal), so the core has to deal only with the common mode signal. So big common inductors in small volumes, that pass big load currents, are possible.

And, with common mode filters, the energy that gets through does not affect other loads across the line (except as common mode interference) but the two power lines act together as a single radiating antenna element. So the energy doesn't go so far before it escapes as radiation. And that radiation (or susceptibility to received radiation) is what most common mode filters are trying to control.

Jeez !

Power line frequency harmonics are typically caused by the short conduction angle of typical capacitor input filters after the rectifier.

EMI filters do not address these 'harmonics' in any significant way *at all*. The most troublesome frequencies are typically the 3rd and 5th harmonics of the power line frequency and effectively completely untouched by an EMI filter.

You clearly haven't tried very hard then !

I doubt they'd want to talk to someone who hasn't even bothered to learn the basics !

The classic EMI filter is there to remove *common mode* noise caused by the switching in the power supply.

Harmonics are dealt with by using 'PFC' techniques.

Graham

I thought as much initially. To suppress that type of interference ( switching noise ) you use a *common-mode* filter.

But you should be aware that 'harmonics' are another issue entirely. See IEC 61000-3-2.

I'm not aware of the use of any differential mode filter in typical EMI applications btw.

Graham

Pooh Bear wrote: (snip)

Then you have missed something.

Any common garden variety EMI filter has both common mode and differential elements. In this typical example of a commercial EMI filter, the common mode (coupled inductor) choke and line to ground capacitors are the (primarily) common mode filter elements while the individual series inductors and line to line capacitors are the (primarily) differential mode filter elements.

(snip)

(quote of first paragraph on this data sheet)

"This series of RFI power line filters has been developed specifically for switching power supplies and is designed to be all the power line filtering needed to control conducted emissions all the way down to 10kHz. High attenuation is provided for both common mode and differential mode interference throughout the frequency range with no degradation of performance due to the large peak currents drawn by switching power supplies."

Did you see somewhere on this data sheet where the filter is "called a common mode filter"?

I didn't refer to any parasitic effects, but to the purposes of specific components, included in this typical line filter.

(Hint: Exclamation points and personal slurs do not improve your arguments.)

If you are refering to commodity power supply AC line filters, then perhaps your impression is due to the fact that common mode choke leakage and loss are also effective on differential mode components, if the appropriate differential mode capacitors are present.

At the actual conversion frequency, however, these small parasitic inductive and resistive terms are seldom sufficiently to allow reasonably sized and priced capacitors to work, without additional discrete differential mode inductors. I'm suprised you've not noticed their presence.

Differential filters are everywhere, in general EMI control of signal lines, on input and output, even if only comprised of a single series resistor, or feedthrough capacitor. The formation of effective common mode filters is increasingly difficult or impractical, as the interconnections involved increase in either quantity or variety of functions.

RL

They are bound to be. That's not their primary function however.

Yeah I do those too ! I don't see a feedthru cap helping PSU emisions much though.

Pls don't 'muddy the waters'. The methods for dealing with classic SMPS noise are well established.

Graham

But it's called a *common-mode* filter you clot !

Parasitic effects are another thing entirely !

Graham

The filter you quote talks about going down to 10kHz. One can infer something about it from that.

The typical 'off the shelf' EMI filter only goes down to 150kHz !

A rather different beast IMHO.

Nor is any filter I've come across been suited to reducing line freq harmonics of order 3 or 5.

I suspect the OP isn't at all familiar with EMI measures

Graham

Differential and harmonic are not synonymous. Harmonic distortion is caused by nonlinear differential loads, but there are non harmonic differential noise problems as well. This example filter is not made for low order harmonic differential noise but for higher frequency, non harmonic, differential noise (like you get when a switch mode supply draws power in pulses at >10kHz from its storage capacitor, causing ripple voltage while the line rectifiers are conducting and connecting that capacitor to the line.

Little doubt of that, since his was a very basic question.