[Prev][Next][Index][Thread]

Re: Variacs in parallel ?



Tesla List wrote:
> 
> Original Poster: "Paul Eugene Kidwell" <tmb-at-ieee-dot-org>
> 
> Hi Bert,
> 
> Bert Hickman wrote:
> >
> > Commercial paralleling chokes are actually small, high current, 1:1
> > transformers. They are simply a single center-tapped high-current
> > winding wound on a silicon-steel core.
> >
> > <snip>
> >                       Vout = [V1-(V1-V2)/2]
> >                                 |
> >                                 |
> > To Wiper of          + Vx/2 -   |    + Vx/2 -          To Wiper of
> >  Variac 1  <---------OOOOOOO----o----0000000--------->   Variac 2
> >            V1        -----------------------        V2
> >                      -----------------------
> >                      -----------------------
> >                           Common Core
> >
> 
> After a lot of thinking, I finally figured out how this would work:
> 
> If the input voltages on both ends are equal, the magnetic field
> in the core would be nil because the two halves of the coil are
> effectively opposite phases and would induce opposite fields that
> would cancel each other. If one side were greater than the other,
> it's field would overpower the field from the other side. Thereby
> reducing the voltage on the greater side, and increasing the voltage
> on the lower side.
> 
> Neat trick :)
> 
> Question 1, how many turns on what size core for a 30 amp system???

I'll give it a shot! Assuming you've got identical variacs, you only
need to handle a maximum of a few volts differential. The core should be
large enough to have at least a couple of square inches of core
cross-sectional area. In order to handle 30 amps, you'll need to use #8
or #10 AWG wire - regular THHN housewire will work. Standard silicon
steel transformer E-I material should be used for the core. A rule of
thumb for the number of turns for this type core material (from Lowden,
"Practical Transformer Design Handbook") follows:

   N = V*5/a  (assumes 60 Hz and 20 kilolines flux density)
Where: 
   N = Number of turns required
   V = Volts across winding
   a = core cros-sectional area in square inches

So... Let's assume you had a core with a cross section of 2 square
inches, and wanted to be able to handle a worst-case wiper-to-wiper
voltage from the variacs of 4 volts. This implies that the total number
of turns would be 10. However, you can add more turns as long as your
winding window is large enough to permit. You don't need a very large
core since you're not developing much total voltage across the
winding.   
> Question 2, How do you tell if it's working??? (I have all the
> stuff I need to start experimenting on this, I just have no clue
> as to how to tell if what I've made is doing it's job)

As long as you've wired it as though it was a single center-tapped
winding, it should work just fine. For testing purposes, you could use a
filament transformer to apply low AC voltage (2-3 volts) across the
winding to verify that you pull only a small amount of current. 

> 
> Question3, you mentioned silicon steel cores, I take it you don't
> mean a standard "E I" type laminated cores??? (If "that's" the
> case, I have something that will fit the bill *perfectly* :)

Regular E-I silicon steel transformer core is just ticket! Remember that
for this application the core doesn't need to be very large.

>SNIP>

> Someone suggested another solution where you wind 3 coils on a single
> core, but there was no further information. Given the choice, I'd
> rather wind a single core, rather than 3, separate ones. Does anybody
> think this would work or have any idea of how many turns on what size
> core???

Usually, independent chokes are used and are the preferred
configuration. While I'm not aware of any simple way to totally
eliminate circulating currents with a single-core approach, the
following should control them. However, it will take a pretty large core
and winding window to handle the large wire guage and multiple windings
required. 

Hook up one end of the chokes to the wipers of variacs 1, 2, and 3 as
shown and connect the other ends together. Center taps are not
necessary, but you should reverse the polarization of one ot the
windings. This arrangement should reduce circulating currents to
manageable levels in an inexpensive fashion. 

            o       
   W1  -----OOOOOO000000------
                             |
                       o     |
   W2  -----OOOOOO000000-----o--- Out
                             |
            o                |  
   W3  -----OOOOOO000000------
 
> 
> I have cleaned off a bunch of cores and have a matched set of 3
> that I could use to do the 3 center tapped method if I have to.
> It's just that I had planned to use these for current limiting
> reactors.
> 

These may be larger than you really need for balancing chokes. You STILL
need to use an external ballast inductance for current limiting, though.

> Any help or clarification would be greatly appreciated.
> Paul

Safe coilin' to you!

-- Bert --
-- 
Bert Hickman
Stoneridge Engineering
http://www.teslamania-dot-com