# Re: ballasting problems

```Original poster: "Bert Hickman by way of Terry Fritz <twftesla-at-uswest-dot-net>" <bert.hickman-at-aquila-dot-net>

Mark,

The Problem:
1. You (and the spreadsheet) assumed that the permeability of the core
(Mu) stayed constant at 5000. This resulted in a ballast which lacked
enough turns. Ferromagnetic materials have a varying permeability - if
you apply too much external flux, they'll saturate, and the resulting
permeability can actually decline to that of air.

2. Because you had too few turns, when you tried to apply a lot of
voltage across the ballast in order to use it as a current limiter, the
magnetizing current climbed much too high, causing the core to saturate.
And, once it began to saturate, the incremental permeability of your
core dropped like a rock.

3. Because of the above, the incremental inductance of your ballast
dropped substantially. At high currents, this may approach that of a
similarly wound air-core coil - providing virtually no current limiting.
Yup! You succeeded in saturating "The Mighty Inductor of Doom" and (at
least before your breaker blew) you were probably pulling hundreds
(possibly thousands) of amps through the winding - it's no wonder the
breaker popped so quickly!

The Solution:
which take into account the maximum, non-saturating, level of flux for
the type core you're using. A typical ballast inductor requires between
125 - 300 turns depending upon core size and desired short-circuit
current. The larger the core area, the fewer turns required _as long as
the core doesn't saturate_. You'll most want to add an air gap between
the E and I sections to prevent saturation and to set the desired
current limit. Once your core starts saturating, current climbs rapidly,
leading to even more saturation, and even more current... well you get
the idea. And so did your circuit breaker (fortunately!).

Your core's cross sectional area is 18". Using this area in the formula
for estimating a core's power handling capability when used as aa
transformer. In the equation below, we'll use a level of 13,000 Gauss as
the maximum flux to prevent core saturation:

P = (A/0.16)^2  (Volt-Amperes)
P ~ 12.6 KVA

Now, lets estimate the number of turns we should use as a function of
the maximum applied voltage. First, let's estimate the volts/turn that
we can support at a level that does not saturate the core:

Volts/Turn = (25.77*f*A*B)/(10^8)

Let:           A = 18 sq. in., f = 60 Hz, and B = 13,000 Gauss

Then:        V/N = 2.21 volts/turn

We want to be able to withstand 240 or 280 volts (assuming you'll run
this off a 0-280 volt variac) so we'll need to calculate the desired
number of turns:

Turns = 280/2.21 ~ 126 turns (worst case for a 280 volt variac)

At 126 turns, you can still have a complete magnetic path (with no air
gap) and the core would not saturate. This is an excellent starting
point, since you can now add a small air gap to reduce the effective
permeability of the core and air gap, thereby decreasing the inductance.
This allows you to set the desired short circuit current. Personally,
with a core this size, I'd go with at least #4 or perhaps even #2 wire
so that this puppy could be used for some future high power work with no
rewinding. By using #4 or even #2 you should be able to go run at 75 A
or 100 A with NO overheating problems and 2-3X that for short runs.

You can also use the tapping factors I showed in an earlier post to
calculate the number of turns for various taps if you wanted to make the
current limiting linearly adjustable for 1 to 5 times Iminimum in 1X
steps. So, if you adjusted the air gap so that Iminimum was 10 A, you'd
end up with taps that would also provide you with 20, 30, 40, and 50 Amp
settings.

Total     Tapping   Tap at    Current
Turns  X  Factor  =  Turn:  (times Imin)
=====     ======    ======   ===========
126       1.00      126         1X
126       0.71       89         2X
126       0.58       73         3X
126       0.50       63         4X
126       0.45       56         5X

Trying to keep Kalamazoo properly ballasted, best regards,

-- Bert --
--
Bert Hickman
Stoneridge Engineering
Email:    bert.hickman-at-aquila-dot-net
Web Site: http://www.teslamania-dot-com

Tesla list wrote:
>
> Original poster: "Mark Broker by way of Terry Fritz
<twftesla-at-uswest-dot-net>" <broker-at-uwplatt.edu>
>
> >
> >
> > Hi Mark....
> >
> > sounds like your inductor either saturated or the H is too small ...
>
> Doubt it saturated....  6 turns on a 18" sq core carrying 40A of juice (fuse
> was rated for 40A).
>
> >
> > you should be looking for around 6 ohms of "resistance" from the inductor
> > to get 40A -at- 240V
>
> That's what my calcs show, too.
>
> >
> > do you have a variac of substantial size ???  ( 30+ amp rating??)
>
> Will tomorrow!  Two, in fact!  (pretty sure they're rated for 240V.....)
>
> >
> > hook up the variac as normal, place the inductor across the outputs of the
> > variac. Slowly apply
> > voltage and record at 10 volt intervals the voltage out of the variac AND
> > the current running thru
> > one of the wires feeding the inductor  ( clamp on amp meter is great
here) .
>
> Didn't think of that method...  duh!  We don't have a clamp-on ammeter,
though
> one should definately be purchased (only \$40, too)
>
> >
> > what I have found in my playing around with scrapped cores is that with
> > 180- 200 winds of #10THHN
> > on various core sizes have resulted in about 2.5 amps per square inch of
> > area ( center leg
> > measurements) before saturation begins to become a problem.
>
> Umm....  I had 6 turns of 6AWG, PVC coated wire on an E-I core measuring
> 3"x6".  According to my handy-dandy speardsheet, that's about 1.8kVA worth of
> ballast.  I guess Chris rewound it with "a lot of turns" (would guestimate 30
> or so).  I don't know if he's tried it yet or not....  By my calcs, that's
> around 700VA....
>
> Even assuming poor quality steel with a mu of 1000 (the books I've seen say
> 3000-5000 for transformer silicon steel), 6 turns still gives me 8kVA.
At 500,
> we start getting into problems (15kVA-ish.).  And we weren't just slightly
> over-current...  that breaker POPPED!
>
> Thanks
>
> Mark
>
> >
> > Scot D
> >
> >
> >
> >
> >
> >

```