### Author Topic: Toroidal core for high power buck converter  (Read 225 times)

0 Members and 1 Guest are viewing this topic.

#### Drirr

• Regular Contributor
• Posts: 54
##### Toroidal core for high power buck converter
« on: June 01, 2020, 06:34:55 pm »
Hi guys,

Im trying to find sufficient toroidal ferrite core for my project. What I need is 82uH inductance. Current 25A, frequency 120, kHz

So.. I found for example http://www.farnell.com/datasheets/2168095.pdf

Material N87, Al = 2880nH

Al=L / N2 >  17 turns is optimal ?

To avoid saturation I want to calculate B = (u0*N*I) /( 2PI*R)  > B = (u0*17*25) /( 2PI* 0,04195m) = 2,026mT ?? only? did I miss something?

I would like to use some of theese cores:
https://cz.farnell.com/w/c/pasivni-komponenty/emc-rfi-potlaceni/ferity-prislusenstvi-feritu/prl/vysledky?st=N87

Thanks a lot,

David

#### TimNJ

• Frequent Contributor
• Posts: 848
• Country:
##### Re: Toroidal core for high power buck converter
« Reply #1 on: June 01, 2020, 07:39:05 pm »
If you want to avoid saturation, you need to look to the core material's datasheet.

490mT @ 20C, 390mT @ 100C. Your peak calculated flux density needs to be less than one of those values, depending on what temperature you estimate it will run at. I usually design for 350mT as the peak flux density to avoid saturation at high temperature, and to account for manufacturing variation between cores.

If your calculated Bmax for the core size you selected is too high, you need to move to a larger core. You can also consider using a different core material with higher Bmax, but this probably won't help if you are way off. Also consider changing the properties of the converter, i.e. adjust the inductance, if you cannot fit a physically larger core.

#### Drirr

• Regular Contributor
• Posts: 54
##### Re: Toroidal core for high power buck converter
« Reply #2 on: June 01, 2020, 07:57:40 pm »

Is there any formula to calculate flux density of toroidal core? I suspect my equation is incorrect, 2mT is too low.

#### TimNJ

• Frequent Contributor
• Posts: 848
• Country:
##### Re: Toroidal core for high power buck converter
« Reply #3 on: June 01, 2020, 08:06:41 pm »
By the way, if you are trying to design this inductor with an un-gapped toroid, you're going to have a hard time using MnZn ferrite like N87. If you need to use a toroid, you'll almost definitely have to look at powdered cores, like powered iron (cheapest), Sendust (slightly more expensive), or High Flux / MPP (higher current, lower loss, more expensve).

This is because the initial permeability (ui) of MnZn ferrite is too high. Remember that Bmax = ueff*uo*Hmax. A powdered core has a lower permeability so it reaches saturation at a higher current. Similarly, you can use a a ferrite core, but it will almost always need to have an air gap.

#### TimNJ

• Frequent Contributor
• Posts: 848
• Country:
##### Re: Toroidal core for high power buck converter
« Reply #4 on: June 01, 2020, 08:10:35 pm »
Bmax = (ueff*uo*N*Ipk)/ lc

ueff = effective permeability
uo = free space permeability
N = turns
Ipk = peak current
lc = mean core length

#### T3sl4co1l

• Super Contributor
• Posts: 15045
• Country:
• Expert, Analog Electronics, PCB Layout, EMC
##### Re: Toroidal core for high power buck converter
« Reply #5 on: June 01, 2020, 11:32:38 pm »
...Or gapped ferrite.  Which sucks in toroids, so you use shapes instead.  A nice RM style perhaps.  Note that the field is intensely divergent around the airgap; you either want to avoid turns in that location, or use very finely stranded wire (Litz cable) to avoid induced eddy currents.

I think you'll find off-the-shelf inductors in that value, and the Bourns 2200LL and related series I believe use Kool-Mu, which has reasonable losses at that frequency.  Design for a ripple fraction under 30% or so and you should be fine.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!

#### TimNJ

• Frequent Contributor
• Posts: 848
• Country:
##### Re: Toroidal core for high power buck converter
« Reply #6 on: June 01, 2020, 11:41:25 pm »
...Or gapped ferrite.  Which sucks in toroids, so you use shapes instead.  A nice RM style perhaps.  Note that the field is intensely divergent around the airgap; you either want to avoid turns in that location, or use very finely stranded wire (Litz cable) to avoid induced eddy currents.

I think you'll find off-the-shelf inductors in that value, and the Bourns 2200LL and related series I believe use Kool-Mu, which has reasonable losses at that frequency.  Design for a ripple fraction under 30% or so and you should be fine.

Tim

Side question, not to diverge from OP too much..but why do gaps suck in toroids? Just from a manufacturing perspective? Off the top of my head, the nice thing about "shaped" cores is that they use bobbins, which means the coil assembly is "modular", core separate from coil. So, if you need to adjust/tune the air gap dimension, no need  to unwind the thing. (Real manufacturers use standard test coils to check if the AL value of the gapped core is correct.)

Is this why? Or are there electrical reasons?

#### T3sl4co1l

• Super Contributor
• Posts: 15045
• Country:
• Expert, Analog Electronics, PCB Layout, EMC
##### Re: Toroidal core for high power buck converter
« Reply #7 on: Yesterday at 01:55:13 pm »
Yeah.  A single gap, has to be sliced in a toroid, and leaves it weak.  You might not be able to wind wire over it, and it might break just from clamping in the winding machine.  If the gap is filled with resin, that's extra steps.  Two gaps, you've just got any old two-piece shape core, except you can't use a bobbin (well, you could slide the pieces into it this time, actually).  And you've got all the disadvantage of the gap's fringing field, and the wasted volume of the bobbin if applicable.

But it does mean distributed-gap materials are ideal, and indeed, we have a huge diversity of those in toroids!

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!

Smf