Off-the-shelf inductor will be too lossy (skin effect?)

[grizedale]

Hello

Please could you tell if an off-the-shelf inductor will be ok for our requirement?

Eg will this inductor be ok….
2209-RC (47uH from Bourns.com)

https://www.bourns.com/data/global/...w.bourns.com/data/global/pdfs/2200_series.pdf

It is the output inductor on a half-bridge smps.

It will see a Max current of following

Average current = 4 Amps.
RMS current = 4.4 Amps.
Peak current = 6.7 Amps

Ripple frequency = 111KHz.

The di is quite high (5.1Amps)
i.e. 5.1 Amps at the switching frequency (111KHz)

We need a high ripple in the inductor to improve our transient response.

Will an off-the shelf inductor with a single winding be ok, or will we need to custom build, with multiple strand conductors to mitigate skin effect?

 

[atripathi]

You need multiple string winding. At 111K skin depth is ~ 4.5E-4 cm. Although datasheet does not indicate any details on winding, and current rating are DC only so it is likely that it is build with single strand, which won't work for rated current at 111K Hz.

 

[mtwieg]

What are your actual requirements for AC losses? I can pretty much guarantee an off the shelf inductor exists, but it might be bigger and more expensive than you wish. But maybe still cheaper than rolling your own solution, especially if you try to use litz wire.

But yeah you'll need the DC rating to be much higher than your expected RMS/average current, since DC ratings don't take core losses or skin effect into account.

Also your link is dead.

 

[FvM]

At 111K skin depth is ~ 4.5E-4 cm

Definetely wrong. You get about 0.2 mm for Cu. So the windings resistance will be increased, but not extreme.

In my opinion, you should use an inductor designed as storage inductor for SMPS, preferably wound with flat wire. The said Bourns type hasn't clear specifications to know e.g. about the core losses at 110 kHz.

 

[grizedale]

Hello

Coilcraft have a skin effect losses calculator.

For my above spec of inductor current, and their single strand PCV-2-473-05L (47u, 5A) inductor, the loss comes to just 43mW.

-amazingly low, considering the di=5.1A and peak = 6.7 Amps., and Fsw = 111KHz

 

[atripathi]

FvM is correct, for 111K, skin depth for cu would be 0.0198399 cm, which would be ~0.2mm. Seems that I punched in some wrong number on calculator.
Sorry about this. Thanks FvM.