Common moide choke current rating

[marginal]

I am writing regarding the common mode choke in smps EMC filter at mains input.

I need a common mode choke of 14.4mH and 2A.

However , i cannot find one which has this high current for this inductance. (at least not one which is of a reasonable size)

This datasheet is typical...

http://www.farnell.com/datasheets/126975.pdf

..the current is never high enough....i wonder if the current quoted in datasheets is common mode or differential mode?

Also, my SMPS is a 35W offline flyback (Flyback PFC [NCP1651] + Flyback PWM [NCP1337])

The avaerage current drawn form the 240V mains is some 400mA...however, the peak is 1.9Amps....therefore i need the common mode choke to handle this 1.9A without saturating.

Since common mode chokes employ "field cancellation", i am surprised that their current capabilities are almost everywhere quoted at such low values.

I would massively apreciate your assistance or thoughts in this matter.

 

[sreid]

You might try to design your own. I think that MicroMetals has design information.

http://www.micrometals.com/index.html

 

[electricuwe]

The current rating of the common mode choke is the RMS-rating for the winding. If you are able to determine the RMS-current properly you can use a common mode choke of significant smaller rating.

Nevertheless 1,9 peak current for 35 W is quite poor (unfortunately you do not state the input voltage range). Add a small differential mode choke to reduce peak (and RMS) current.

 

[sreid]

Checkout Coilcraft and J. W. Miller.

http://www.coilcraft.com/

http://www.bourns.com/bourns_jwmiller_cobranding.html

 

[sreid]

Can you use a smaller filter capacitor after the bridge rectifier? More B+ ripple but a lower peak current. Current mode PWM controllers don't care much about B+ ripple as the current loop attenuates 120 (100) Hertz very effectively.

 

[marginal]

hello,
thankyou all for getting back.

I apologise my rms input current is actually 0.271 Amps...(which would have a peak of 400mA if it was sinusoidal)

I have rigged up the NCP1651 in discontinuous mode, and to keep losses down, i set it up for a longer secondary conduction time.....which unfortunatley gives the 1.9A peak primary current which is drawn from the 240V mains via the EMC filter.


I am not using any capacitance after the bridge rectifier as i want it to be nice half sines there. I have seen some app notes putting capacitance there, though i am keen to avoid it.

 

[Skogsgurra]

You are OK then.

The common mode choke laughs at the peak current. If, and you really have to check that, your RMS current is below the winding's thermal rating, which it seems to be, then your peak current is OK.

The common mode choke is what its name says: Common Mode. That means that you can run any current through it as long as you take the same current back again in the other winding. It will not saturate as long as the two windings carry same, but opposite currents. Think bifilar winding.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[electricuwe]

Strange design for a SPMS without smoothing cap.

-This means 2*line frequency ripple on the output voltage.
-operating thr Flyback only for a few cycles during peak voltage condition --> high stress, poor efficiency.

 

[Renovator1]

electricuwe said:

Strange design for a SPMS without smoothing cap

Sounds like a boost PFC converter, actually.... Yep, just looked up the datasheet for the part, it's a PFC controller chip.

 

[electricuwe]

Yes, no (large) cap is o.k. for a PFC. But is the PFC works properly peak current in that application should be:

- line: 35 W *sqrt(2) /(efficiency *240 V)
- PFC switch: a little twice that value

both much below 1.9 A.

Is the PFC-choke properly rated ?

 

[GonzaloEE]

I've seen CM chokes in my job usually work with current pulses of some 3 times as much the nominal current. These pulses last only 3ms, so having 3A peaks is completely normal as long as the input meets the rated RMS current of the choke.

Adding Y1/Y2 caps from lines to GND may help further reducing the CM ripple, as most CM chokes I've worked with, show self resonance at freqs. 1 decade below the caps.

 

[gabisuciu]

try to find at EPCOS C.M choke.
The inductance is important, but you have to look at the characteristic L to frequency. Usually for EMC you need high induct. at 1-50MHz.
You need also a film capacitor after the bridge for the H.F switching current.