Author Options:

Magnetic Levitation Tram? Answered

Hello there instuctables,
I need your help with a project i am attempting, i am in the process of making a maglev tram and i have a couple questions i need help with?

1. I have a simple circuit for a linear motor that i understand to the most part but i need help with finding an electromagnet to use / make, it needs to run on batteries and obviously the main concern here is weight and stability so the smaller the better but still needs to be strong enough to propel the tram along the track so i suppose my real question is what ammount of winding, resistance and core material would be suitable for whatever voltage i could provide!

2. Next i suppose what way round should i place the magnets, should i place many electromagnets on the rail with sensorsand permanent magnets on the tram, or should i have a powered electromagnet on the tram and permanent magnets on the track?

3. Also what hall sensor should i use to detect the magnetic fields, unipolar switches or ratiometric sensors?

4. And finally control, how would i control the motion of the tram i.e. speed and direction?

Id like to thank you all in advance for any help you could provide and i hope my description is accurate enough!!
many regards,

Photo of simple linear motor circuit (img.1)
When the hall sensor (switch) senses a pole of a magnet, it activates the transistor, in turn magnetising the relay and causing the current in the electromagnet (motor) to run one direction, then as the opposite pole is detected the relay de-magnetises and reverses the current thus reversing the pole on the electromagnet! This is designed so that with permanent magnets on the track in an alternating fashion of poles (nsnsns) the tram is either pushed or pulled along the track! 


Beg, steal, or borrow "An Engineer in Wonderland", by Professor Eric Laithwaite, the pioneer of maglev.

Laithwaite's books are horrifically expensive now, ( he is dead).

However the Christmas lectures he gave based on the book are available on Youtube.

these lectures seem to be all about gyroscopes! very interesting stuff, but not really related, there are a few on maglev but nothing particularly in depth?

Full size or model?

A linear motor is essentially a 3 or more phase motor opened out flat.

You need the phased to get moving it reacts with the aluminium track to produce eddy currents which repel the motor giving both traction and lift.

Your not in the right area with your diagram.

I assume a model so why not lift with permanent magnets and control with a raised track side and small wheels.

For an easy drive put a prop on the back, this was suggested many times in the past a soon as aircraft became popular.

Otherwise you going to need to design a sort of flat stepper motor or synchronous and some kind of drive for it.

what about this new design i am uploading? its a three phase motor essentially, the psu will be connected before the Vac and before this will be an inverter before both to convert dc batteries to ac voltage!


I am not too sure you understand how the linear motor works. The field has to move, i.e. cycle through the coils. at present your not showing how this is achieved.

just connecting single phase AC to a set of coils isn't going to do it.

Its not single phase anymore? At the coils there is firstly a capacitor which, i believe, alters the phase to -90', then an unaltered coil, effectively 0' (although all coils affect phase but in this explination ill just say the coils have no effect), and followed by a inductor on the last coil, making it +90'! Surely this would cause one electromagnet to be reppelling constantly one attracting constantly and one holding? Or have i missed the point? Or is it just a bad drawing that doesnt explain it properly?

No, you'd need to do a circuit analysis to prove it, but take it from me, it won't work.

okay well what about an r-l-c network across the centre magnet, using a variable resistor to keep it balanced, then as the resistance changes it pushes the centre magnet out of phase, thus changing the balance?? does that sound more along the right lines or am i completely barking up the wrong tree?

Okay well can you give me some reasoning why? Or possibly some idea on how to make it work?

Read Laithwaite's books. I can't think of a clearer, more intuitive guide, to understanding the concept of TRAVELLING, or ROTATING magnetic fields.

Basically, you need to perform a vector analysis of the components of the fields in your coils. You will find your proposals yield only static fields, in the sense they possess no rotational content, merely varying in amplitude.

Okay so if i understand correctly.... The three phases need to travel through all electromagnets in sequence i.e phase 1 = coil 1 as phase 2 = coil 2 as phase 3= coil 3, then phase 1 = coil 2 as phase 2 = coil 3 as phase 3 = coil 1.....etc etc? ? So all phases push through all coils in sequence? If that makes sense i will watch those youtube vids with great interest!

Yes, you're creating a TRAVELLING wave, not a STANDING wave, in other words, one that satisfies the wave equation - which I can't write in here.

Okay im sure i can find the equation on wiki or the such! So if i were to say use a darlington array to fire the magnets as a stepper motor would am i now along the right track, no pun intended??

im thinking the digital inputs are contolled by hall sensors in dc steppers?


Okay got a slightly better idea, i think, how about a time pulse from a 555 timer into a Jk flip flop into a stepper driver controlled by logic?

Only for a very, very light load. Basically you're making a coil gun.

Oh i see so really the best route is the 3 phase inductitrack and forget dc entirely?

Okay so i suppose the next logical step is to ask the best way to create 3 phases from a single supply?

By far the easiest way would be to buy a cheap single- three phase motor driver.

Think i found a couple on ebay, one is a 1hp IMO idrive single to 220v 3phase or theres Omron 230v sysdrive.... What one should i be getting?

oh and forget lift for a minute as this will be done seperately

I suggest a lot more reading. This scheme won't work,