intentially design high High proximity losses in Buck output inductor

Dec 12, 2017 8 Replies

Playing with a 150VDC buck design that requires 4mH inductance. 600mA output current. I'm designing it on a 2616 pot core in Intusoft Magnetics designer software. It telling me set the AL to ~ 450nh and ~85 turns. Its running at 150KHz.



DC resistance is less than 1 ohm but AC resistance is something like 15 ohms.



The heat generated is not a problem but would the high AC resistance act like a snubber of sorts?



Its estimated to have a loss at full load of about 400mW. this is tolerable.



If excellent efficiency it not of highest concern is there a benefit to leaving the inductor with high AC winding resistance losses.


The alternative is of course miltifilar, litz, or foil winding to bring the Rac and Rdc values closer together.



thoughts?


Coilcraft! Winding inductors is a last resort.

John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com

Cant do coilcraft on this one. I'm going to chicken out and have it wound miltifiler but was wonding about the advantages of high Ac loss in terms of ringing reduction. in my application the reduction or elemination of the diode subber would be beneficial.

There aren't many places where a mediocre Q-factor is useful.

Certain powdered irons (namely #26 and #52) are too lossy to use in most, say, flyback converter applications. But they're not so lossy that they just go "thud", so you still need a damper or snubber to absorb the ring-down EMI.

Actually, it's pretty close, I've done it before. Sooo lossy...

Anyway, that's the same as hooking a big fat resistor across the winding. It's not a high-frequency-only thing, like an R+C damper. It's just... lossy.

An application where efficiency doesn't matter, like my recent current limiting circuit breaker device, benefits from losses to the extent that:

  1. It doesn't cut into accuracy of the current limit (a solid-iron core would be dominantly resistive, and not really do any filtering anymore);
  2. Any loss in the core is loss I don't have to handle elsewhere (in transistors, resistors, TVSs, etc.). More components handling more power = more cost. Anywhere I can economize is a win.

This is a good application for something like a #26 powdered iron, instead of a gapped ferrite core.

If you're using a pot core, it sounds like you're already well above the point where anything like this can apply.

So... maybe, sometimes -- but no?

Also, on the subject of both #26 toroids and 150V buck converters, you may be interested to see this?

formatting link
First tried a #26 here, cooked itself. Then a Kool-Mu, also cooked itself (but slower). Finally changed to gapped ferrite, just because I have a shitton of 'em. Now it runs quite cool, because, well, of course it does...

Tim

Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: https://www.seventransistorlabs.com/ "mook Jonhon" wrote in message news:2jGXB.15850$lD5.9612@fx39.iad... > Playing with a 150VDC buck design that requires 4mH inductance. 600mA > output current. I'm designing it on a 2616 pot core in Intusoft > Magnetics designer software. It telling me set the AL to ~ 450nh and > ~85 turns. Its running at 150KHz. > > DC resistance is less than 1 ohm but AC resistance is something like 15 > ohms. > > The heat generated is not a problem but would the high AC resistance > act like a snubber of sorts? > > Its estimated to have a loss at full load of about 400mW. this is > tolerable. > > If excellent efficiency it not of highest concern is there a benefit to > leaving the inductor with high AC winding resistance losses. > > > The alternative is of course miltifilar, litz, or foil winding to bring > the Rac and Rdc values closer together. > > thoughts? > > >

I've burned the paint off MicroMetals powdered-iron cores in simple

100-amp buck switchers. Cool-mu fixed that.

But I'd guess that using inductor losses to serve as a snubber isn't going to work. But one could spend a lot of time finding out.

Right, as long as the resistor doesn't fry. A series cap can function only to reduce the dissipation.

John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com

If you want to go non-linear, the market of ready-made inductors instantly becomes virtually empty. Maybe the

3R1 saturable chokes for the PC PSUs remain...

And if even that is not enough and you dream of multi-aperture devices, then man, you have a serious problem...

Best regards, Piotr

Coilcraft and Bourns and Coiltronics and Sumida and Bussman and a zillion other people make thousands of inductors and transformers. And they do custom designs, with the engineering usually free. And you can put parts in series or parallel. It would take a pretty exotic application to require you to wind your own inductor.

The standard parts will go nonlinear, too.

Some people enjoy winding wire on cores, but it seldom makes economic sense.

John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com

OK, if you let them wind your custom inductors, then there is no question about it. I thought you mean using only the standard parts.

Selom a good idea. If you can afford going fully custom, then just go.

Best regards, Piotr

I do that 98% of the time.

Why so?

John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required