>> Note that this means the core needs to store 4.4J. =A0You can play all t=
he
>> games you want with turns, wire thickness, core material, etc, but there
> is
>> no way to escape this basic physics.

In my case, the inductor can fully saturate and still do its job. I
now realize that it's all about the core material. The coil only has
to keep the rate of current rise below a threshold. I was originally
looking for a way to model inductors way past their saturation point.

SCRs are available in many di/dt ratings. However, devices rated for
1000A/us are usually priced around $500 each. Commodity SCRs that I
checked out had ratings of 100-200A/us (at 125C junction temp). I'm
trying to work out the best way to build this circuit. Expensive
silicon devices and forgo the di/dt hassles or cheap easily
replaceable SCRs with some not-so-easy inductor designing.

> However/also, it seems to me that an air wound core will also be achievab=
le
> that works OK.
> This page contains an apparently superb calculator that goes beyond the
> simplistic formula that I offered.
>
> =A0 =A0 =A0 =A0 =A0 =A0 =A0http://hamwaves.com/antennas/inductance.html
>
> How good it is is uncertain, and what the target R is is unknown to me, b=
ut
> it seems that 4 to 6 turns of eg 15mm silver plated copper tube wound in a
> (tight ) 100 mm dia spiral with minimum inter turn clearances will produce
> inductances of the order required. Resistances appropriateness need check=
ing
> but are probably bearable.
> They give what seems to be AC resistance (ie skin effect modified
> resistance) but I haven't even tried to check for reasonableness.
>
> Plugging in a few figures I get the following.
> Use courier or other fixed width font to view:
>
> Dia =3D coil mean diamyter mm
> t =3D conductor thickness (copper tube)
> W - width or coil length mm
> N =A0turns
> Rac - resistance impedance at 100 kHz. =A0????
>
> N =A0Dia =A0 t =A0 W =A0LuH =A0 Rac
>
> 4 =A0100 =A010 =A050 =A01.5 =A0.007
> 4 =A0100 =A015 =A070 =A01.14 .005
> 4 =A0200 =A015 =A070 =A03.53 .01
> 6 =A0100 =A015 =A0100 2 =A0 =A0.008
>
> Bifilar winding using eg silver plated 1/8" copper tubing will reduce AC
> resistance due to skin effect reductions.
>
> Values are such as to suggest that a handful of small gauge piping, some
> silver plating (not expensive) and some playing, should yield an acceptab=
le
> result.
>

I was hoping to have something below 5mOhms, or at least below
10mOhms. There are other areas I can give or take from to reduce the
effective series resistance. Lowering series resistance would free up
cheaper SCR choices.

A coil made like you described will probably work quite well, and it's
quite feasible to build. I'm still convinced I can do the job with a
smaller coil (if a balance of saturation vs inductance loss can be
found). However, to do that, looks like it will be all
experimentation.

The wikipedia has some good sections on magnetism which mention some
exotic materials like mumetal.
http://en.wikipedia.org/wiki/Permeability_(electromagnetism)

I found some good information from MIT too:
http://web.mit.edu/6.013_book/www/chapter9/9.4.html

Many thanks to the list and all the great minds that inhabit it.
Russell has helped me see an obvious solution that I missed, of a
really big coil.

-- =

http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist