[Prev][Next][Index][Thread]
Re: transformer design
Tesla List skrev:
>
> Original Poster: bob golding <yubba-at-clara-dot-net>
>
> Hi all,
I`ve been kind of waiting for a question like this.....
>
> Went to my favorite scrap yard today and hauled out a 3 phase
choke. It is
> marked "3 phase reactor 112 amps 415 volts" When I got it home I quickly
> removed the windings,3 layers of 1/8"x1/4" section cotton covered copper
> wire. This has left me with a core of the following dimensions 10" x10" 3"
> with two windows 6" x 3" x 2".
This means that you have a stack, 3" high, with with 3 legs and 2
windows, each 2 " wide, Right? Great find: You will be able to pull 5kVA
out of this one.
I have looked in the archives and think I
> can work out a design. What worrys me is the permeability of the core as it
> was a choke rather than a transformer,
Hakuna Matata, dont worry. If the laminations are thin, and that means
if they are 3 lams per millimetre, they are M6 oriented
laminations=state of the art. A choke or a transformer is the same
thing, basically.
> and whether I should interleave the
> laminations or leave them as they are in an E and I configuration and
> suffer the core losses,
If they are stuck together well, with transformer varnish, by all means
leave them like that. If you are handy with a file, you can file the
mating surfaces, Lapping the surface with valve grinding compound on a
thick glass plate will doo wonderfully too, or take the thing to a
machineshop, and have the mating surfaces ground. You have 200 £`s worth
of lams, so a bit of spending is in order.
Core losses are a result of the lamination grade, not any airgap. And
even if you interleave the lams, there will be an airgap, depending upon
how well you pack the stack.
> or try to design them out by increasing the turns
> per volt.
Turns per volt, that`s the question. I would say 1,25 volts per turn.
Now for the winding layout: If you wind the primary in the central leg,
and the secondarys on the outside legs, you will have half the
flux-density on the outside legs, and I am not sure that would be any
good, so while you are down at the machineshop, ask them to saw the one
leg, and one window off the core, leaving you with a U-I core.
The window area is a bit small, but by running the copper at 5 amps/sqr.
mm^2 you will be allright.
You should now go to my site, and wach the homemade pig page, to get an
idea of how the windings are layd out.
Http://home5.inet.tele.dk/f-hammer/tesla/tesla.htm
Make 2 primary bobbins, 40 mm long out of 4 mm thich material, and wind
88 turns from 2,5mm wire, in 7 layers. Use 0,1 mm mylar as interturn
insulation, that makes for a neater coil.
Make the coils identical!. Then connect the bottom turns together, and
apply 220-240Volts to the outher turns.
The secondaries are assuming that you want 15000 volta -at- 400mA (get
prepared to spend a 80nF cap!)
This coil should be layer wound on a bobbin without endplates, and with
0,05 mm mylar as interturn insulation. Don`t skimp here, paper may do,
but still, mylar is the thing to use, to avoid further ponderings about,
wether the thing will last.
Wire diametre will be 0.5mm. Coil length= 110 mm. Start by winding 5 mm
from the ends of the insulation, progressively making the spacing
bigger, untill the top turns are 15 mm from the ends. (this will happen
easier than expected, it is really hard to wind out to the same length
as the layer below, without the insulation flipping down, and the turns
sliding out. Make 36 layers in all, for 6545 turns.
There should be 4 mm btwn the bobbins when the thing goes together, fill
this out with scrappings of mylar, sink the thing into an oilbath, and
ramp it up with a variac. Make a jacob ladder and enjoy!
Hope this helps, Finn Hammer
>
> bob golding