HVLabs.com

"MagLev" Project

Electro-magnet suspension of a permanent Magnet

This circuit design actually comes from Rick Hoadley's Magnetics website, which details how this circuit works in great detail. If you are interested in constructing one of these, I suggest you print and read through his work.

If you've ever tried to levitate a permanent magnet using an electro-magnet before, you would have found that there was no equalibrium and that you could not get the magnet to stay floating in free air. ie, It would always try to jump up to electro-magnet or fall out the air.

This circuit pretty much creates an equalibrium between the two and allows the magnet to hover. The circuit is by no means the first to achieve this, it was invented and demonstated a long time ago (search google). There is a comercial toy availble fairly cheaply that uses this principle, but there's no fun to just buying something is there!

Basically this circuit works by taking two reference inputs (hall effect sensors) which measure where the magnet is and how strong the pull on the magnet is (electro-magnet) and subtracts the two to create an error signal. The error signal controls the current in the electro-magnet either making it stronger or weaker. So really what the electro-magnet does is switch on and off very quickly to create just the right amount of pull vs gravity.

Parts:

Many many bits, no not really :-)

- Go to Rick's site and get the circuit diagrams and read them thoroughly.

- Parts, 3x TL084 (faster response) IC's, 2x 555's, 6x 10K 10-18T cermets (v.resistors!), 2x SS495 (ratiometric) Hall effect sensors (must be ratiometric, I got that one wrong at 1st!), 1x Electro-Magnet (Build your own, Iron/steel core, plastic form and some copper enamalled wire), 1x Neodyme Rare earth magnet (fairly big), 1x IRF644 (or any N-fet) and a bunch of resistors and caps.

As for the electro-magnet construction, I tried two different coils before I got it right! So here's mine that worked. (Hand built by a good friend of mine in his spare time).

The details for the coil I used (Center pic): Wire is stranded copper of a 3mm gauge. Coil resistance at cold is 2.4ohm, supply voltage is 5v, current is 2A. Or if I want it more powerful I use 12v, current is then 5A (things start to get hot at this voltage, so I mainly use 5v.

Right pic: The power comes from an old PC power supply that gives you the necissary +12v, 5v, -12v and Grnd. You could use regulators and a straight 12v supply, but why bother!

Circuit Tuning Tips:

Rick Hoadley's site gives you a good walk through for doing this, but I have found I needed to change a couple of things to get mine to work. Such as changing the 555 timer frequency to 4Khz (3.3nF cap) to make it a faster response (and less whine noise). Also I added a Led on the circuit output to give me an indication when the electro-magnet was on and by how much (dims led). Another thing worth adding is a high speed power diode across the electro-magnet to absorb any RF and back EMF.

As for setting the pots up, I found that putting them close to the center (5K) and working around that areas helped me find the correct setting. I also found you need to adjust the position pot alot at the start to stop the magnet getting too close to the electro-magnet. Setting the Duty-cycle-limit was easy as you just use your scope and find the maximum and then work back from there.

I found that the magnet would sometimes oscillate up and down getting gradually bigger until the maget would either fly up and get stuck on the magnet or drop. This only happened after 5 mins or so of the magnet being suspended. Weight can be used to dampen it slightly or what I've found to be the best thing is using a strip of thick aluminium (a dielectric) to resist change and stabilise the object. I have taped 4 3mm strips of aluminium together to form a block and that works very nicely. It can be above the magnet or below.

Outcome:

It does take alot of time and patience tuning the circuit to work (about 1 week on and off!) but when it does take the magnet straight out of your hand with some invisible force and hold it floating in mid air for the first time, you will be very pleased (I was!). You can attach anything you like to the magnet and have it floating without strings attached!

My Levitator:

So far I have managed to levitate a CD, an aluminium fan blade a small figurine (an apache Indian I think, can't find my toy soldiers!) a steel bolt and a paper UFO. You can just have the magnet free floating. I have used two sizes of magnets in the pics, but have now stuck with the large 1" Neodyme for more lifting power.

Click to Enlarge pics


Apache Indian figure on small Neodyme with 1st coil	The control board, built on veroboard, fairly neat hey!	Control board + power transistor	Power Mosfet (IRF644) on Cpu heatsink (doesn't get hot tho)


Aluminium fan blade with steel bolt and Large neodyme. Able to spin this with air blown from small cpu fan. Looks excellent!	Same again from different angle. This will float like this in stationary position for well over an hour.	The pulling power is quite extraordinary even at 5v, it will try to stay in the center even if pushed!	Large Neodyme on it's own with aluminium block under neath for dampening. Can be taken away and will float for about 10 mins before oscillations occur. Tune with Position pot.

Finished MagLev, Support structure made from MDF.

Videos:

A short video demonstrating my paper UFO, flying!

Right Click and choose save as.

WMV 6Mb

Lately I have been trying to spin the suspended object up to high speeds using a small motor and a large plastic/rubber wheel (to keep it away from the elctro-magnet), It has been quite sucessful as theres is almost no friction. I do wonder if you could use a high voltage source to push the object around with the Ions, although I believe it may interfere with the sensors.

Another thought is that you could possibly power something suspended by creating another coil on the magnet to capture power as this is half of a regular transformer as it is.