inductor sizing

Hi,

For a filter inductor in an LC filter, does the inductor size (core volume) scale linearly with inductance * current rating^2, all other variables the same, ie. core material, frequency etc.

Also I need some toroid cores for 200kHz LC filters, 250uH and 50Amps peak current, are these going to be huge cores and what core materials would be best to use, ferrite or powder cores?

I've downloaded the

formatting link
"Magnetics Inductor Design Using Powder Cores" program:

formatting link

cheers, Jamie

Reply to
Jamie Morken
Loading thread data ...

That's stored energy, which should be proportional to core volume.

Iron powder cores tend to burn their paint off in situations like this. Permalloy powder is very good but expensive. The "Kool-Mu" cores are almost as good, but cheaper. Gapped ferrites are OK here, too.

John

Reply to
John Larkin

Huhhh..dunno about that 1st question. I'm no magnetics expect...but will take question #2 for $2 million..

Let's say you're filtering a DC supply with an LC filter. You have 50 amps DC passing through the 250uH inductor.

By Ampere's law. H= 0.4*Pi*N*I/Le CGS units Le = mid circumference of the core N = turns I = amps

The field intensity H * ur = B flux in gauss ur = relative permeability (Note: ur varies, ~2000 for some ferrites)

If B>>Bmax. core spec (around 4000 gauss for some ferrites) ...that inductor acts like a paper weight (resistor + an air core L).

My guess (without doing the math) is that you'll DC saturate a ferrite and will end up choosing a powdered core.

As for powdered size, the Magnetics Inc program might select a size you like.

D from BC British Columbia Canada.

Reply to
D from BC

--
ISTM that, after the spec\'s were defined and the questions were:
"are these going to be huge cores and what core materials would be
 Click to see the full signature
Reply to
John Fields

Get a life, Fields.

John

Reply to
John Larkin

Oops..have to fix a few things..

H=(0.4*Pi*N*I)/Le

And... ur decreases especially on the knee to saturation.

Like biasing a transistor, best not to over bias. There has to be useful permeability for your target inductance. I've seen some app sheets suggest operating a ferrite toroid at

1/2Bmax.

D from BC British Columbia Canada.

Reply to
D from BC

Why do you need that much inductance? 250 uH at 200 KHz is an impedance of 300 ohms, which seems a lot for a 50 amp circuit.

John

Reply to
John Larkin

--
There once was a fellow named Larkin
who thought that to him all should hearken.
 Click to see the full signature
Reply to
John Fields

JF is a poet but doesn't know it ;-)

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

Stick to designing 555 circuits; your career as a poet doesn't look promising.

John

Reply to
John Larkin

To properly design that inductor more info is needed, such as; Iac (For core, proximity and skin effect losses), Idc (For I^2R losses), Ipk (for core max flux density) and E*T (core saturation) along with the operating frequency already given. Max ambient temp, inductor temp rise and size constrains would be helpful. Given that we can then offer an opinion. Cheers, Harry

Reply to
Harry Dellamano

--
his feet show it, they\'re long fellows. ;)

Remember that one from grade school? :-)
Reply to
John Fields

:-) My question here would be something along the lines of "why on earth do you need 250uH at 200kHz and 50 amps??" To me it implies either a huge amount of power (i^2*L*f/2; at rather high voltage), or a poorly/strangely designed filter system. If it's lots of power, you should of course expect to be using relatively large parts, and it's worth thinking very carefully about the design of the inductor and the system into which it fits--probably not something to just jump into without experience at lower power first. -- I wonder what it is that Jamie is actually trying to accomplish.

Cheers, Tom

Reply to
Tom Bruhns

Yep ;-)

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

Hi,

The inductor is for generating 120VAC at 60Hz:

formatting link

I have a simulator of the PWM controller for this circuit, if I set it to output 3.6kW at 120VAC, with a 250uH inductor and 10uF cap I get this waveform: (THD + noise = 1.7%)

formatting link

With an inductor RMS current of 28.5amps and peak current of 45Amps and capacitor RMS current of 1.8Amps and peak of 5.2Amps.

If I change the inductor to 50uH and the cap to 100uF, I get this waveform: (THD + noise = 2.4%)

formatting link

With an inductor RMS current of 32amps and peak current of 86Amps and capacitor RMS current of 17Amps and peak of 46Amps.

This larger 100uF cap is required with the smaller inductor or else there is way too much ripple. Also this 100uF cap will be expensive as it has a large current rating requirement.

It seems like a tradeoff, if going for the 250uH/45Amp inductor, what toroid material would be good and still economical to use?

I am planning on ordering some various toroids so I can experiment with winding them and stacking them if necessary, so am looking for the best core types to be using for this.

cheers, Jamie

Reply to
Jamie Morken

F type ferrite from Magnetics Inc starts into saturation around 1/2 oersteds. You can't even get 1 turn on it with 30 amps DC!

0.5 = (0.4*Pi*1*30)/le

The math gets horrid.. A 75 centimeter magnetic pathlength... Ewww :(

However, I'm just learning this stuff myself.. Maybe there's still some useful inductance in saturation.

D from BC British Columbia Canada.

Reply to
D from BC

--
The injured knight screams.
The lance has pierced his armor.
 Click to see the full signature
Reply to
John Fields

I saw Beowulf last night..cool graphics, Jolie is hot but I did wayyy too much fast forwarding.

Here's my version of Beowulf :)

King makes noisy party. Next door neighbor Grendal gets pissed, shows up, eats people. Grendal doesn't eat the King even when taunted. Mystery. Hired gun Beowulf shows up. Beowulf gets to boink the kings wife if he kills Grendal. Beowulf kills Grendal. Beowulf boinks kings wife. Grendal's mom gets pissed, shows up, kills people. Beowulf visits Grendal's mom.. What a hot milf! The deal. Beowulf boinks Grendal's mom and her magic makes Beowulf 'Super King'. (Grendal's mom likes popping out human/monster hybrid kids.) All's well until some jerk steals the horn from Grendal's mom. Her new son goes postal. Beowulf kills that one too. Beowulf dies. Successor to the throne wants some xxx with Grendal's mom. What a hot milf! And the cycle continues..

D from BC British Columbia Canada.

Reply to
D from BC

formatting link

formatting link

formatting link

Jamie, Your Vout + PWM waveforms confuse me. The frequency of the output and PWM is about 0.35Hz. There is no PWM just a square wave and the filtered output is shifted 0 or 360d and at that amplitude and phase cannot be the filtered output. Please advise, Harry

Reply to
Harry Dellamano

?? Seems to me a reasonable starting point would be about 5uH and

10uF (comprising perhaps ten 1.0uF or twenty 0.47uF high quality polypropylene caps) as a first stage out of the FET bridge, followed by perhaps 1uH and another 10uF (which doesn't have nearly as high a ripple current), to yield a fraction of a volt p-p 200kHz ripple at the output, assuming a 3 ohm load and 150V out, input switching to +200 for 4.3us and -200 for balance of the 5usec cycle. Looks like it does OK through the range of duty cycles needed to generate the output 60Hz sinewave. Though the first-stage capacitor has to handle a lot of ripple current, I'll bet you'll find that doing it this way, with a couple stages, will be cheaper by far than a 250uH inductor that meets your needs. I believe Wima MKP10 caps would do OK in this application. If you have frequencies lower than 200kHz on the output, I'd recommend solving that with a better control system.

Perhaps I don't understand just what you are doing, but at least the way that's obvious to me to generate a "clean" 60Hz wave from the partial schematic you provided suggests to me that you don't need anywhere near 250uH to do what you want.

Cheers, Tom

Reply to
Tom Bruhns

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.