Tour : click picture below for previous

Back to Mike's Electric Stuff

Click picture below for next

smhifi.gif (937 bytes)smallmyst.jpg (1332 bytes)More fun with neodymium magnets.

Neodymium Iron Boron (NIB) magnets are extremely powerful, and allow effects to be seen that aren't possible with normal iron or ferrite magnets. I discovered the wonders of these magnets at Bill Beatty's site, which you are strongly advised to read first for background info and more fun experiments.

I remembered that I already had some of these which I had previously removed from scrap hard disks. Most hard disks with voice-coil actuators contain these magnets - look for drives greater than about 50MB, which have no motor visible on the outside of the disk housing. When you dismantle the head actuator, you will usually find one or sometimes two magnets fixed to flat metal plates. They are are either lightly glued, or located by pegs and held in by their own magnetism. If glued, the easiest removal method is to bend the plate - the magnets are very hard and will usually snap off the plate instead of snapping in half!

Neodymium magnets can also be found in certain "in-the-ear" type headphones and upmarket in-car speakers.

I've just found that small cylindrical NIB magnets made by Assemtech are available cheaply from They are hidden in the 'proximity sensors' section of the catalogue, and not listed in the index. They are also stocked by UK distributor Rapid Electronics. I suspect these are the type used in small headphones.

Being cylindrical, these can also be used for 'slow roll' experiments, e.g. rolling them between the fins of an aluminium heatsink.

Size dia x height, mm Flux (gauss at surface) Assemtech part no. Farnell Stock no. & 1-off price (GBP, ex VAT) Rapid Stock No. & 1-off price (GBP, ex VAT)
3 x 1 1500 M1219-1 723-0345 0.50 78-1064 0.40
3 x 2 2500 M1219-2 723-0357 0.53 78-1066 0.43
4 x 3 3200 M1219-3 723-0369 0.67 78-1068 0.53
6 x 2 2500 M1219-4 723-0370 0.67 78-1070 0.50
10 x 5 3500 M1219-5 723-0382 0.67 78-1072 0.68

For larger magnets in the UK, Eclipse Magnetics, who do a range of neodymium magnets from approx 3.00 for larger magnets ( 24mm dia 20mm thick 4500 gauss).  If you want to buy a LOT of Neodymium magnets cheap, try this Chinese manufacturer.

Thanks to Tony Dixon for this info : (May 2003) For Neodymium magnets I can also recommend Magnetic Applications Ltd,  Berkhamsted, Hertfordshire. Their 4x1mm mags were 100 for 20 (+ post & VAT) and they were helpful friendly and took a credit card order over the phone.

Below are a few more NIB magnet experiments.  More neodymium stuff at,

Jumping Magnets

Lay a flat NIB magnet over a straight wire on a flat surface - fix the wire down with tape. Connect a 1000 uF capacitor (the RIGHT way round!) to the output of a 12-24v power supply, current limited to about 1-2A.
When you connect the capacitor to the wire, the magnet jumps up and slides off the wire (if you get the polarity right).
The capacitor provides an initial surge to lift the magnet up and overcome friction, and the DC supply provides a constant force to slide the magnet away from the wire.

The magnet needs to be placed slightly off-centre, with the wire about two-thirds along its length. Reverse the polarity or turn the magnet over if it slides towards the centre.

                 /       /|
                /       / |
               /       /  /
  /===========/       /  /=======\
  |          /       /  /        |
  |         /_______/  /         |
  |         |       | /          |
  |         |_______|/           |
  |                              |
  |                              |
  |                              |
  |      1000 uF                 |
  |          _       pushbutton  |
  |       | | |+     ____        |
  *-------| | |---*--O  O--------/
  |       | |_|   |
  |               |
  |               |
      - 12-24v +
       1-2 amps

Alternatively, lay a thin wire (e.g. wire-wrap or magnet wire) over the magnet - it jerks upwards (or downwards) very rapidly.

Self-shaping wire

Lay a 4" length of thin magnet wire (0.2mm, 32 AWG) or a single thin strand from a stranded wire along the length of the magnet, and tape the ends down an inch from each end of the magnet, leaving enough slack to pull the wire into an arc about 0.25" above the magnet. The wire will pull down very rapidly, bending itself into a Z shape when current is connected. Then, reverse the polarity and it bends to a Z the other way. Set the PSU current limit to < 0.5A (or use a lamp in series), otherwise the wire will get rather hot! Pictures below show 'before' and 'after'.

magwire1.jpg (3525 bytes)magwire2.jpg (3595 bytes)

Shimmying wire

Using the setup above, connect the wire to the output of a small (<6VA) 6-9v transformer (small transformers will self-limit the current so the wire doesn't melt), and watch the wire dance about as you move the magnet around.

Slow fall 2

magfall.jpg (4290 bytes)The copper tube slow-fall demos described at Bill Beatty's site work quite well with  the flat magnets obtained from hard disk drives, but the gap between the magnet and the pipe wall reduces the effect. Stacking several magnets to make a square-section improves things a lot however.

A very nice braking demo with flat NIB magnets can be done with a long (6"+) chunky (3mm+ thick fins) aluminium heatsink, by dropping the magnet between the fins. This has the advantage over pipe that you can see the magnet falling. If the gap is big enough, fit 2 magnets together so the magnet-to-fin gap is very small (~1mm)- fall rates of about 5-6 seconds per foot can be observed. If the fins aren't chunky enough, or the gap is the wrong size, stand two thick-base heatsinks on their ends, back to back (shown left), with a gap slightly bigger than the thickness of the magnet, & drop the magnet into the gap.

Hard disk maglev demo

What to do with the rest of the drive when you've swiped the magnet...

Take a scrap 5.25" hard disk, but one which still spins when you power it up (This may work with 3.5" drives, but the magnet tends to be attracted towards the motor shaft, so 5.25" drives allow the magnet to be further away). Remove the heads and actuator assembly (you've already done this to get the magnet, right?). Verify that the disk still spins up with the head assembly removed. You may find that it spins down after a few seconds as it can't read any data from the disk (well you'd have trouble reading too if someone had ripped your head off!).

Remove the platters and spacers from the motor shaft, and re-assemble them so that all the platters are at the top of the shaft, with all the spacers at the bottom, giving a nice thick stack of aluminium. Tape a neodyminum magnet to the end of a strip of plastic. The reson for using a strip, rather than, say, threads, is that the magnet tends to pull towards the centre and also tumble over - the strip constrains the magnet's movement to upwards only. Fix the strip to one of the side holes of the drive chassis, so the magnet rests on the disc surface, near the outer edge, as shown below.

When the drive is powered up, the magnet slowly levitates to about half an inch above the disc surface. Pressing the magnet down slows the disc rotation substantially.

This might make a nice demo of how hard disc heads fly, by putting a magnet inside a model of a read/write head.

maglev.jpg (11911 bytes)

    An interesting note sent to me on a practical application of neodymium magnets : I was reading your section on neodymium magnets, and remembered something from my summer at Cedar Point (an amusement park in the midwestern US). We have a rollercoaster called Millenium Force. It is probably the most high-tech 'coaster out there. Believe it or not, it uses neodymium magnets mounted along the tracks before the station and aluminium fins on the trains as a braking system! It proves to be very effective - it takes a 3 ton train going 65 MPH and stops it smoothly using 24 1 FOOT LONG neodymium magnets...

Diamagnetic levitation

wpeA.jpg (43793 bytes)

Some materials, including bismuth and graphite exhibit an interesting property called diamagnetism. This is effectively the opposite of being magnetic - they repel magnetic fields. Unlike the repulsion effect you get whith the like poles of two magnets, these materials repel both magnetic poles equally, and this makes it possible to achieve a stable freestanding levitation effect.
Unfortunately the diamagnetic repulsion is a rather weak effect, so some tricks are needed to get levitation. The above picture shows the setup I used to make a small neodymium magnet hang in free space. There is a large Neodymium magnet at the top, which applies a lifting force to a smaller magnet, which is placed in a slot in a block of diamagnetic graphite. The large magnet provides the bulk of the lift against gravity, the graphite acts as a stabiliser, repelling the small magnet if it gets too near either side of the slot. This setup needs extremely fine adjustment to balance the lifting force against gravity, and I used a micrometer screw slider assembly scavanged from an old fibre-optic fusion splicer. The black stuff under the top magnet support is soft rubber foam to stop the small magnet shattering if it gets dislodged and attracted towards the top magnet.

I got the graphite I used from Wondermagnet.  Also check out their really cool magnetic field viewing film.
You might find that the graphite used in electric motor brushes will work for this experiment.
Lots more diamagnetic experiments here and here and here.

smhifi.gif (937 bytes)smallmyst.jpg (1332 bytes)

Click picture above for Previous

Back to Mike's Electric Stuff

Click picture above for next