[TCML] More on balacing chokes from Bert Hickman

```The three choke configuration looks like it is for a 4 stack. My bad.

<paste>

Designing/Using home-brew balancing chokes:
> ================================
> Suppose you want to use chokes, but not spend \$122 on them? What should we
> plan for? Let's look at the 1256D in detail to get a clue. The wound core
> is about 12" in diameter, and along the outside it has about 8 turns/inch
> (TPI) of #10 AWG wound around all but 4.25" of the periphery. So the total
> outer periphery of the core that's covered by a winding is about (Pi*12 -
> turns or so on the total winding. If the unit is configured for 0-280 volt
> operation, the variac develops around 1.05 volts/turn.
>
> If we can limit the wiper variation between variacs to no more than +/-1
> turn, then the worst-case voltage difference between variacs would be
> 2.1 volts, but if we can only guarantee +/-2 turns, then this will rise to
> about 4.2 volts. So, worst case, each paralleling choke must be capable of
> handling at least the maximum expected output current, and the core size
> and number of turns on the chokes must be sufficient to support at least
> 2.1 volts across each half of the choke. The paralleling chokes can simply
> be center-tapped windings wrapped around a common core. If we make two
> chokes using wire sufficient to handle 30 amps, and a third choke capable
> of handling 60 amps, then we can actually get by using only three chokes
> for four variacs as shown in the diagram below. Not that making these
> chokes is MUCH easier than making ballast inductors - fewer turns are
> needed on a much smaller core.
>
> Let's plug in some nunbers... we can use the standard transformer equation
> to determine the volts/turn for a given core area and operating frequency
> (from Eric Lowdon - "Practical Transformer Design Handbook", 2nd ed., Tab
> Books, 1989, 389pp - a great book!):
>
>    V/N = 4FfaB*1e-8 = volts/turn
>
> Where:
>    V = Volts
>    N = Number of Turnn
>    F = 1.1 (waveshape factor for a sine wave)
>    f = 60 Hertz
>    a = core cross sectional area (in square inches)
>    B = Maxwells (or lines)/square inch = 80,000
>
> Then, for a core with a cross section of 1 square inch:
>    V/N = 4*1.1*60*1*80,000*1e-8 = 0.213 volts/turn
>
> We need our chokes to be able to handle 2.1 volts on each leg, so:
>    N = 2.1/0.213 = 9.85 turns - use 10 turns on each "leg" of the choke.
>
> Let's allow 800 circular mils (CM)/ampere for wire sizing for intermittent
> duty - at 30 amps, this means we need about 24,000 CM or about #6 AWG. In
> order to handle twice the current we can use two #6 AWG wires wound in
> parallel or a single #3 AWG winding. If you used a larger core area, you
> could get by with fewer turns. The cores from small junk NST's, small
> variacs, or even taped cores from surplus current transformers should be
>
>
>                                 |----> Vout = [V1-(V1-V2)/2]
>                                 |
>                                 |
> To Wiper of            N turns  |  N turns            To Wiper of
>  Variac 1  <-----------OOOOOOO--o--0000000--------->   Variac 2
>            V1          -------------------        V2
>                        -------------------
>                        -------------------
>
>            I1
>   W1  <-----------
>                  O
>           Choke  O   I1+I2
>            #1    o-----------
>                  0          |
>            I2    O          |
>   W2  <-----------          O
>                      Choke  O  I1+I2+I3+I4
>                        #3   o--------------> Vout
>                             O
>            I3               O
>   W3  <-----------          |
>                  O          |
>          Choke   O   I3+I4  |
>            #2    o-----------
>                  0
>            I4    O
>   W4  <-----------
>
>
> Hope this helps save you some \$\$\$... \$490 for Superior's technically
> correct answer is probably too painful for most coilers... :^)
>
> -- Bert --

_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla
```