Filter Chokes

[Leslie Horn]

For a power supply I want to build, I need a pair of 300 millihenry chokes, at 7 amps, and with a DCR of less than .2 ohms.

After using several search engines, I have come to the conclusion that I should wind my own. The cores I tried, EI 96 (center column 32mm x 80mm) is too small, as the required 700 turns of wire will not fit in the winding window. I know that a larger core will require fewer turns, but the DCR will not decrease any if at all, due to the larger turns.
I want to use the smallest core possible. Also, the wire must be fairly large (three #12 AWG, trifiliar), to keep the DCR low. This requires a larger core, simply because of the physical window dimensions required for the copper wire.

I'm baffled. I only want to buy the core laminations once.

Failing that, where can I pick up a pair of 300mH, 7A coils???

[Sch3mat1c]

#12 is good for 10A, why the heck would you need trifilar? Making a 2V 100A supply or something?

TIm

[Magura]

300mH and 0.2 ohm sounds like an oxymoron to me......

Well, sure do-able but will turn out rather big and expensive.

Take a look at one of the inductor calculators and think it over.


Magura

[Leslie Horn]

Originally posted by Sch3mat1c
#12 is good for 10A, why the heck would you need trifilar? Making a 2V 100A supply or something?

TIm

I want to keep the DC resistance low. This will be a choke-input +/- 60 Volt linear supply at 7 amps. The power supply will have a minimum load of a few hundred milliamps, and I want fairly good regulation and efficiency. I could add a shunt regulator, I suppose, but then there goes the good efficiency of a choke-input supply. I'd rather just keep the DC drop in the chokes comparable to (preferably less than) the diode drops. I calc'd out the resistance of the wire, and it is surprisingly high in smaller gauges, with the required length.

[Chris Jennings]

Hi.

Are you sure you need 300mA at 7 A ?

My initial calculations suggest you will need about 50kg (110lb) of core material for each (using M6 laminations) I've not calculated the size or stack, since they would need to be custom lams.

For 1 ohm resistance you would need about 22kg (44lb) of core.

Chris

[Magura]

That matches what I estimated. Now to get to the practical part, it is virtually impossible to wind an inductor with such wire gauge, even when running severel 2mm wires. My experience is that 2*2mm wires are about where it turns into a real raw deal.

Magura

[Leslie Horn]

Originally posted by Chris Jennings
Hi.

Are you sure you need 300mA at 7 A ?

My initial calculations suggest you will need about 50kg (110lb) of core material for each (using M6 laminations) I've not calculated the size or stack, since they would need to be custom lams.

For 1 ohm resistance you would need about 22kg (44lb) of core.

Chris

Ouch. I was afraid of that.

But then again, 7A is the max, and it can swing. The critical inductance is about 150 mH at a couple of hundred ma... I could eliminate the air gap and let it swing...

But capacitor input is starting to look more and more feasible

[Sch3mat1c]

LMAO! n1 Chris.

So let's see. 7A through 0.2 ohms produces a voltage drop of 1.4V (dissipating 10W), a loss of 2% regulation. So uh...if you can't even measure that...why are you concerned about it in the first place?

Oh, and WTF is it with people and SHUNT REGULATORS? Out ye foul vermin!!!!!!!

Tim

[AndrewT]

Hi,
a choke regulated supply must be designed for a narrow range of current. There is a formula somewhere.
If you fall outside the necessary current limits (both hi & lo) then the choke stops working for you.
You are talking about 0.3A to 7A over 20:1.
The numbers supplied by our contributors show why hi current chokes are VERY expensive and also confirms why choke regulated supplies are generally found in lo current applications e.g. valve amps.

[Circlotron]

Chokes 4 everyone!