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

Re: Ballast choke questions



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

Weazle and all,

Nice catch on the core! I've been meaning to follow up on my earlier
posting, since some off-list work with another coiler has resulted in 
modifications to the tapping table. First, answers to your questions 
are interspersed below...

Tesla list wrote:
> 
> Original poster: "J. B. Weazle McCreath by way of Terry Fritz
<twftesla-at-uswest-dot-net>" <weazle-at-hurontel.on.ca>
> 
> Hello Coilers,
> 
> I've spent the last two evenings disassembling a large transformer
> that I acquired with the intent to rewind it as a ballast choke to
> use with a 5 KVA pole pig I'm getting.
> 
> The stack of E and I laminations measures 3.5 inches thick, with
> external dimentions of 7 inches by 7.75 inches.  The center most
> leg, or the middle of the E measures 3.5 inches by 2.25 inches.
> The two "windows" measure 5 inches by 1.5 inches.
> 
> Now for the questions to you transformer and choke experts:
> 
> 1) How much power should a core of that size be capable of?

You have a GREAT core for a ballast choke! A first-cut estimate of power 
handling capability for a silicon steel core with a core area of "A" 
(in square inches) is as follows:

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

Solving for your core with A = 3.5*2.25 = 7.875 square inches:

     P ~ 2400 Volt-Amperes

BTW, another useful relationship is the volts/turn as a function of core
size:

Let:
      A = Cross sectional area of center leg in square inches
      f = Operating Frequency
      B = field (in Gauss) - use 13,000 (most silicon steel)
  
   Then:
     Est Volts/Turn = (25.77*f*A*B)/(10^8)

   Using your core:
     V/T = 25.77*60*3.5*2.25/10^8 = 1.58 volts/turn

So, if you wanted to be able to withstand a full 280 volts without
saturating the core you'd need to have about 280/1.58 = 177 turns. 

> 
> 2) When re-assembling the core, should the laminations be
>    put back interleaved as originally found, or should all
>    of the E's be put together facing the same way?

Align all the E's and I's, since you're doing to add a gap in the magnetic
path between the E's and the I's to prevent saturation of the core.

> 
> 3) If all E's are aligned, should there be an insulator put
>    into the gap between the E's and the I's?

Yes. This can be virtually any non-magnetic non-compressible material.  It
shouldn't take much of a gap (10's of mils at most) to have a significant
impact on the maximum inductance, but this gap is essential to prevent core
saturation. 

> 
> 4) I'm going to put multiple taps on the choke windings for
>    selecting different currents.  Should the first 79 turns
>    as per the table below go nearest to the core, or on the
>    outside of the windings?

Be careful: my previous post assumed that you could freely "fit" 250 turns
on your ballast inductor. Unfortunately, this is NOT the case for your E
core - although your winding window is relatively large, you'll need to use
about #8 AWG to handle 40-50 amps, or #10 AWG to handle 30-35 amps for
reasonable run times. If you use #8 THHN, you'll only be able to fit
between 4 and 4.5 turns/inch, and with a 5" long window, you'll only be
able to fit about 21 turns/layer, and you'll only be able to fit about 5
layers. This will limit you to about 100-105 turns for #8AWG. 

If you used #10 AWG instead, you could fit about 25 turns/layer and 7
layers for a total of about 175 turns. For a variety of reasons, I'd
recommend using the latter for a 5 KVA system. Regarding Richie's comments
on using taps, they could indeed reduce available winding window and they
add complexity to its construction. However, if you locate them along the
"outside" areas of the winding the taps should have a minimal impact
reducing the available wiring area between the core legs. And, if you use
flat copper strips wrapped by heat shrink tubing to bring out the taps, the
impact can be further minimized. Note that this configuration is close to
the value predicted above for 280 volt operation WITHOUT Saturation with NO
air gap (the approach recommended by Richie in his previous posting).

The table I previously provided resulted in linear inductance changes per
tap over a 10:1 range in 10% increments. However, most coilers probably are
more interested in obtaining linear increases in short-circuit ballast
current/tap instead. Since the calculation for short circuit current
limiting places the inductance in the denominator (I = Vin*2*pi*F/L)) the
table needed to be modified to make the current adjustment/tap linear.

I've modified the turns factors to reflect this, and have also reduced the 
range to 5:1 between maximum and minimum current. You can simply use four
strategically placed taps to obtain this 5:1 spread. The total number of
turns will be constrained by the window area and size of wire (per above).
By adjusting the core gap, you can directly control the level of minimum
current level while using all turns.  The short-circuit current at various
taps will then be 2X, 3X, 4X or 5X the minimum level. In the table below,
the number of turns for each tap were calculated by multiplying the total
number of turns by the "multiplying Factor" in the table. I've completed
this for a ballast with 105 turns (#8 AWG) or 175 turns (#10 AWG).  

                 Multiplying  #8AWG  #10AWG   Short-Ckt    
     % Total L:    Factor       N       N      Current
        100%        1.00       105     175      1*Imin 
         50%        0.71       100     124      2*Imin  
         33%        0.58        94     101      3*Imin 
         25%        0.50        88      88      4*Imin 
         20%        0.45        81      78      5*Imin 

<SNIP>

> Any and all comments are welcomed.  I want to do this once and do
> it right the first time.  Thanks in advance.
> 
> 
> 73, Weazle, VE3EAR/VE3WZL
> 
> Listening: 147.030+ and 442.075+
> E-mail:    weazle-at-hurontel.on.ca
>            or ve3ear-at-rac.ca
> Web site:  www.hurontel.on.ca/~weazle

This was probably "clear as mud" but hope it helped a bit!

Best regards,

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