Choke Inductors in SMPS

[AlienCircuits]

Hello, be prepared for a lot of naive questions.

I'd like to use a "choke" inductor as my conversion inductor (as opposed to EMI filter) in a boost PFC design. I've never made a boost PFC converter before, so I am unsure if I'm making the right choice with this inductor.

My PFC will take line voltage from 70 to 140 VAC and convert it to 210 VDC @ 26 watts. I wanted to minimize ripple current, and so I tried to find as big of an inductor as I could in the right price/current rating/environment range.

I found this:
https://www.epcos.com/inf/30/db/ind_2008/b82721a_j_k.pdf
Part number B82721A2122N020 (6.8mH 1.2A per winding)



Now, my confusion is that this is a choke inductor. Does this imply that it is not designed to be in the converter process, but just as a common mode choke for EMI? If I connect the 2 windings in parallel, I will have 3.4mH with a 2.4A rating, but then I am unsure if the common mode configuration will have negative effects on the 100kHz switched current into it. What if I connect them in series? To increase my buying power, would using the same part for both my EMI choke filter and my conversion inductor be feasible?

 

[schmitt trigger]

No;

EMI inductors cannot carry and DC bias, they saturate if you do, and its inductance goes off the cliff.

You require a proper inductor which can sustain DC bias. This means either a powder-iron core (which is acceptable for 100 Khz) or a gapped ferrite core for higher frequencies.

Now, I understand your desire for low current ripple. But PFC inductors cannot be chosen arbitrarily.

What you require to do is select a controller from one of the major semiconductor vendors. In the datasheet, they will have all the equations required to design the converter (which depends on your specs) and choose the proper components. Among them will be the value and current rating of the boost inductor.
If you are lucky, the vendor may even have some app notes and software to assist you in the design.

Then you can go the the Epcos, Coilcraft, Murata, Vishay or other magnetic component vendor and choose the closest commercial value that fits your needs.

EDIT: you will still require the inductor that you have shown in the datasheet, but for its intended purpose: reducing EMI. But there is more than an inductor for a good EMI filter, you will also require some X/Y capacitors.

- - - Updated - - -

The UCC28019A from Texas Instruments appears to be a simple to use, low cost device which could easily fit your requirements.

 

[AlienCircuits]

Thanks for the reply.

I did not arbitrarily choose my inductance value and specs. I have been using datasheets, app notes, and text references to double check, triple check, etc. that I am choosing the right inductance and current characteristics for what I want to do. I will have about a .38A DC current with a .1A ripple at full load and minimum input voltage with an inductance value of 4.7mH. Changing from 4.7mH to 3.4mH or 6.8mH will affect the ripple current, but nothing so out of range that it will invalidate the other design choices of the MOSFET, diode, and filter cap.

My confusion is mainly on the use of choke inductors. From what you said, I need something called a PFC inductor and I do not want to use an EMI choke. Well, how do I know when a choke is an EMI choke vs. a PFC inductor? I've seen the inductor referred to as both a choke and an inductor in the references I'm reading, so its always confusing.

Does the word choke imply EMI application? If not, how would I know that the particular inductor I linked in the datasheet is an EMI choke as opposed to some other type?

 

[FvM]

The main problem is not with EMI chokes as such but with the specific "current-compensated" type. it means that the specified inductance is not available for the load curent.

 

[AlienCircuits]

The main problem is not with EMI chokes as such but with the specific "current-compensated" type.

So this sounds contradictory to what schmitt trigger commented about chokes (other than the current-compensation thing, which I will take as advice from both of you to not use this particular choke part).

 

[FvM]

In don't see the statement contradictictory. It's a good advice to look for PFC inductors, because major manufactors have dedicated products for this application.

But other inductors can be used too, generally all types that are intended as an energy storage inductor, e.g. for buck or boost converters. I was specially referring to the current compensated type. If you consider it's operation principle, you understand why it can't work as a energy storage inductor.

 

[schmitt trigger]

read my post carefully, I said "proper inductors which can sustain a DC bias".

DC bias saturates magnetic materials. Inductors applications which require to sustain DC, like a PFC converter, either have a discrete air-gap ground into the core, or have a distributed air-gap built in the magnetic material itself.

Such inductors will clearly have a spec which indicates: xxx uH @ yyy DC amps.

 

[mtwieg]

The term "choke" is used for different applications. In the most general sense it refers to an inductor whose purpose is to pass primarily DC current. However, another important distinction is whether the choke is meant to store energy (like those used in SMPS converter stages), or whether it's used for suppressing AC current (common mode filter chokes, DC biasing chokes, etc). So don't dwell on the "choke" terminology, just find an inductor rated for your desired current, frequency, and inductance.