[TCML] charging reactors

[jimlux]

bartb wrote:
> You certainly don't want to use THHN sized insulations and a single 
> layer would not do.

I was just thinking of that as a "bound" to physical size of an air core 
inductor.  If the 500 foot spool were, say, 10 mH, then a similar sized 
spool of magnet wire with 4 times as many turns (obviously wound 
differently for voltage reasons) would be 160mH or there abouts.



  Magnet wire is the preferred choice here. The
> voltage is distributed through the winding and you want to attain the 
> inductance in the least amount of "space" with the minimum voltage 
> stress across the worst case opposing layers. A single layer is not 
> possible without getting ridiculous with geometry.

Yes.


> 
> Everyone seems to be thinking in terms of heavy gauge and heavy 
> insulation wire. If I remember Jim's specs, it was 10kv to 20kv in the 
> 500mA to 1A range at about 5H. This really isn't that difficult to 
> achieve. For example, consider a 1"D x 4.3"L core. Now imagine 100 
> layers of 30Awg magnet wire. The result is a 5H inductor good for 1A and 
> the worst case stress across the layers is 400V with 20kV applied. 
> Obviously, this wire size is small, DCR is 166 ohms and net impedance is 
> 1.88k ohms. You could zap this wire size with too much current if no 
> current margins were taken into consideration. So maybe head up to 2"D x 
> 4.75"L core and use 24 Awg (robust). This inductor ends up an outside of 
> 6.5"W x 4.75"L. Pretty hefty inductor, but it could be wound on a bobbin 
> and would certainly do the job. Because the layers are still 100, the 
> voltage stress is still near 400V worst case across joined windings.

I'd worry about surface creep across the end of the winding.  Multiple 
windings in series to distribute the voltage along the length of the 
winding seems a good idea.

> 
> Anyway, there are many layer/winding possibilities, but magnet wire is 
> what you want here. Going with a thick insulation (if thinking about 
> voltage standoff) is not practical nor is it needed in this case. I know 
> Richie Burnett wound his large ballast with thick insulated wire, but 
> the inductance was in the low mH range (it's a ballast application). 

That's sort of what I was wondering.  (without actually just running the 
equation with some guesses as to the size of wire, or *gasp* walking out 
to the garage and measuring it...)

> Quite a different cookie than what would be needed for Jim's need.
> 
> Take care,
> Bart
>