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

BIG magnets available





---------- Forwarded message ----------
Date: Sun, 5 Oct 1997 12:53:48 -0500
From: Gary Johnson <gjohnson-at-kansas-dot-net>
To: tesla-at-pupman-dot-com
Subject: BIG magnets available

There have been several comments and questions about the availability of
Neodymium Iron Boron magnets on the list lately, so I will mention that I
have a few that I could sell.

Several years ago I came into possession of 320 pounds of unmagnetized NIB
blocks, left over from some Delco-Remy research project.  The blocks were in
two sizes.  The larger was about 6.2 by 6.2 by 2.4 cm, weight about 700 gm.
Three of these blocks were glued together.  The smaller were about 5.5 by
5.5 by 2.2 cm with a mass of about 500 gm, and had four blocks glued
together.  The blocks could be chiseled apart with some difficulty.  I did
not think about heating the blocks to the melting point of the glue, as has
been suggested, and may try that next.

I looked for a magnetizing system within reasonable driving distance of
Manhattan, KS and could not find one, so decided to build one.  The minimum
magnetizing force was over two million A/m, so after a few iterations, I
decided to build a system capable of 6000 A in a 100 turn coil just large
enough to hold the blocks.  The material magnetizes very quickly, so only a
short pulse is necessary.  But 100 turns of wire has enough inductance that
the pulse cannot be too short.  

The basic idea of a magnetizer is a capacitor bank, a coil to hold the NIB
block, and a switch.  For the capacitor bank, I used 96 capacitors rated at
a nominal 1400 uF and 450 VDC each.  The capacitors were placed in six rows
of 16 capacitors each.  The conductors were sections of square aluminum
tubing, 1.75" square, 0.125" wall thickness.  Each bus would carry at least
1000 A so mechanical strength was important, as well as low resistance.  The
large cross section also helps reduce inductance.

I did not know how to build a mechanical switch that would handle 6000 A
multiple times without excessive erosion and losses. (Still don't)  So I
used two Westinghouse 2N3890 SCRs in parallel.  These have a steady state
rating of between 250 and 300 A, and a surge current rating of 4500 A.  

The coil was made of eight layers of 12 ga. magnet wire on a square former.
Both ends of each layer were brought out so layers could be paralleled.  The
first layer had 100 turns, the second had 99 turns, etc.  The increase in
diameter from one layer to the next caused the inductance to increase.  The
fewer number of turns caused the increase to not be as large.  As has often
been pointed out on this list, multilayered coils have about the same
inductance as a single layer, just lower resistance.  The inner four layers
in parallel had a resistance of 0.38 ohm and an inductance of 258 uH.  The
outer four layers had a resistance of 0.42 ohm and an inductance of 284 uH.
The currents would not split exactly in half, but close enough for my purposes.

This was an unusual design project in that it went exactly according to
Circuit Theory I.  Calculations before construction and measurements after
construction were surprisingly close.  We put a 20 ft cord on the SCR gates
so we could be behind a sturdy bench at the first firing, but no explosions!
Magnetic forces drive the two parallel conductors together near the coil
with a satisfying thump, but that is about all.  The coil tries to get
shorter and fatter with each firing, hence tends to eject pieces of potting
compound after a while.

Anyhow, I now have rows of magnets stuck to the walls of my metal building,
probably more than I could ever use!  I checked the price of magnets in
Edmund Scientific, and choked!  They are asking $245 for a 2" by 2" by 0.5"
block (5.08 cm by 5.08 cm by 1.27 cm).  I can beat that price by a factor of
ten on a dollars per gm basis and still be overpriced.  Of course, the main
problem of selling the magnetized versions is that I have to find a BIG box
and pack the magnet at the center with rigid styrofoam.  (UPS might get
concerned if a package jumped up and stuck to the side of the truck.)

I can guarantee that these magnets will pinch your fingers and magnetize
your screwdrivers.  They will magnetize your monitor so it display a rainbow
for ever.  I passed one around class once and a student wiped it across his
student ID card, which did not work the next time he used it.  Usually EE
students are brighter than that, but this kid played tuba in the marching
band.  You don't have to be crazy to play that position, but it helps.  You
cannot demagnetize these units electrically.  The only way is to heat them
above the Curie temperature.  I do NOT claim that these magnets are actually
useful for anything.

The claimed internal magnetic flux density is about 1.1 Tesla.  The highest
I actually measured external to the magnet was about 0.8 T.  The magnetic
moment for the 6.2 by 6.2 by 2.4 blocks is about 82 Am^2, and three times
that for the case of three blocks glued together.  The material is
anisotropic and can only be polarized in one direction.  If one draws a
diagonal across the large face of a block, and then rotates the line by 12
degrees to align with a small dent in the face of the block, one is close to
the direction of the field.  That is, the flux basically comes out of a corner.

Anyhow, if someone wants a magnet, send a check to:

Johnson Energy Corporation
Box 1032
Manhattan, KS 66505

Single block 6.2 by 6.2 by 2.4 cm      $40
Triple block 6.2 by 6.2 by 7.2 cm      $80

The price includes shipping and handling anywhere in the 48 states by UPS
ground.  If you live elsewhere, send me e-mail and we will explore
possibilities.  If you are driving down I-70 in the Manhattan area, let me
know a day or two in advance and you can pick one up (at a substantial
discount).

Gary Johnson