# Solenoid Artistic Function - Magnet and Glass Tube

My brother is a glass blower and he mentioned an interesting idea to me, which I believe is similar to a solenoid. The idea is to seal a glass tube with a magnet inside. Then wrap the glass tube with a copper wire from end to end. The copper wire will be spaced so that you can still see into the glass to see that magnet. On each end of the wire there would be an electrical connection. Then by reversing the electrical polarity the magnet can be seen floating back and forth. What is the best approach for flipping polarity and circuit design with the Arduino Uno? How much voltage and/or current do you think would be required for a 1 foot tube 2 inches wide, with a strong magnet from magnetic key box you'd stick under your car, and copper wire 2 mil thick?

Not sure that 'floating' is an accurate description since the magnet would be resting on the glass and (hopefully) sliding from end to end as you power the coil.

Impossible to say how much current/voltage the coil will require, partly because the magnetic flux needed to move the magnet is unknown and partly because you can trade off current versus number of turns to produce that flux - the voltage requirements would then be determined by the required current and the resistance of the required number of turns. However, the wire you're mentioning sounds far too thick unless you are planning a coil to take massive current. You would probably be much better off designing it for a few thousand turns of much finer wire, but how many turns, how much current is needed with those turns, and how thick the wire needs to be to support that much current, you will need to find by trial and error.

PeterH:
Not sure that 'floating' is an accurate description since the magnet would be resting on the glass and (hopefully) sliding from end to end as you power the coil.

Impossible to say how much current/voltage the coil will require, partly because the magnetic flux needed to move the magnet is unknown and partly because you can trade off current versus number of turns to produce that flux - the voltage requirements would then be determined by the required current and the resistance of the required number of turns. However, the wire you're mentioning sounds far too thick unless you are planning a coil to take massive current. You would probably be much better off designing it for a few thousand turns of much finer wire, but how many turns, how much current is needed with those turns, and how thick the wire needs to be to support that much current, you will need to find by trial and error.

Ok, very good. Please explain more about the resistance of the required number of turns in the coil. I don't quite understand that. How would you know. Isn't it dangerous to send electricity across a bare copper wire?

Not sure if it will work. If you power the solenoid it will suck the magnet in to the middle if you power it the other way, it will try to turn the magnet, not really pushing it to an end. On the other hand, if you only wrap the bottom half with wire, you can use different current to repel the magnet so you may suspend the magnet at different vertical heights.

Doesn't have to be "bare" copper wire - can be magnet wire, which has an insulating layer of varnish or equivalent over it.

Google it - lots of choices.

Biggest missing element if you want the arduino to be able to control the position of the internal magnet is that you need a feedback measurement method to know where the magnet is at any given time so the controls know how much and in what direction to power the electromagnet(s). A software PID controller would have to be set up most likely. While the overall effect would be very cool and appear simple the complexity of control is much more difficult then what one might expect. And on being able to reverse the polarity of a electromagnet requires an H-drive much like used for DC motors where you want to be able to change the direction (polarity) as well as speed (amount of average power).

Lefty

for a 1 foot tube 2 inches wide, with a strong magnet from magnetic key box you'd stick under your car

With these dimensions you have no chance to get it to work. Even if you could get enough magnetic field the magnet when it moved would break the end of the tube. What is likely to happen is that the magnet would not move along the tube but just flip over.

liudr:
Not sure if it will work. If you power the solenoid it will suck the magnet in to the middle if you power it the other way, it will try to turn the magnet, not really pushing it to an end.

I don't think so. When the coil polarity reverses the force will reverse as the magnet poles are still in the same position. The round tube will not allow the magnet to rotate to reverse its poles. Think about why swolenoids work. If the magnet and coil poles are opposite the thing will fly to the closest end as that will be the dominant attracting force and the closer it gets the stronger the force.\

The end breaking problem by the magnet when hit can be solved by using a small magnet covered in 2 layers of soft material on both flat sides.

The feedback signal for the control loop I'm still thinking on that one...

liudr:
Not sure if it will work. If you power the solenoid it will suck the magnet in to the middle if you power it the other way, it will try to turn the magnet, not really pushing it to an end.

I don't think so. When the coil polarity reverses the force will reverse as the magnet poles are still in the same position. The round tube will not allow the magnet to rotate to reverse its poles. Think about why swolenoids work. If the magnet and coil poles are opposite the thing will fly to the closest end as that will be the dominant attracting force and the closer it gets the stronger the force.\

The end breaking problem by the magnet when hit can be solved by using a small magnet covered in 2 layers of soft material on both flat sides.

The feedback signal for the control loop I'm still thinking on that one...

You did this before? Any proofs beyond your reasoning that it would work? What is the expression of magnetic force on a magnetic dipole?

Maybe the feedback siganl can be solved like this. Just maybe. Let' see what they find wrong with it and maybe we can improve it.

You can't touch the magnet with anything as it will affect the position and look awful; but..

You can place a round piece of reflecting something in top of the magnet thing. At the end of the tube you palce an LED and an LDR (light depending resistor) then you get the reflected light from the mirror and convert it to voltage with the LDR. The closer the magnet the more reflection you'll get and the higer the voltage. You will need to use infrared or something to avoid ambient light from affecting the thing.

Whatever you do with the magnet it has to move freely and you may need a hole in the middle or something to allow the air flow back and forth otherwise the piston efect it will create will slow it down or prevent it from moving altogether when the air gets compresssed. You can also vaccum the tube after done and seal it again as they did it the past with vaccum tubes. Vaccum tubes again? =(

You did this before? Any proofs beyond your reasoning that it would work? What is the expression of magnetic force on a magnetic dipole?

I haven't mentioned specific quatities or anything.. It's just simple reasoning, You just need to look at solenoids woking to realize about that. The cores don't rotate.

And you? have you done that?. Do you have poof of that other than your reasoning? What's the equation for magnetic force depending on the distance and current? Did you calculated that? With what values because n0 one here has nmentioned anything like specific numbers for currents etc. This is just brainstorming.

This is just brainstorming.

That is one word for it but it is another word beginning with b and a lot shorter.

You want to see this magnet and yet you want a coil around it, this means spacing between the turns, and so reduces the magnetic field.
Solenoids have hundreds if not thousands of turns and are not two inches in diameter.

All of a sudden your magnet goes from

a strong magnet from magnetic key box you'd stick under your car

To one that cannot rotate in a two inch diameter tube, .... that does not add up.
I am not sure what you are going to do with a feedback sensor for the magnet's position, what are you going to do with the information?

Forget the glass tube just try it in a cardboard tube, I bet it won't even move with the current you can put in it.

And no I haven't done the maths, I just have had many years of playing with large magnets.

I did not posted this therefore I never mentioned any dimensions of anything. The author was. If he wants to make the magnet "float" it needs to be about the same diameter to avoid spinning. By the way I wasn't answering to your post. I think I know now why Grumpy is in your name. Maybe a similar word beginning with J will describe you be better.

encryptor:
Ok, very good. Please explain more about the resistance of the required number of turns in the coil. I don't quite understand that. How would you know. Isn't it dangerous to send electricity across a bare copper wire?

Rather than me try to explain the theory of electromagnetism to you, I suggest you go do some research for yourself to learn about the theory behind electromagnets, and then look for practical examples of people who have made solenoids similar to the one you're envisaging.

very true. Ok will do. I believe I've heard of solenoid cannons that can shot out a magnetic rod from it. This does sound like a high energy project b/c I want the magnet to float. I think the PIR detector set just right and timing to adjust polarity would work and use of magnetic wire. Ok a H-bridge for polarity switching. I like the idea about a coated magnet to prevent glass from breaking or hearing the "ting" noise constantly. Ok a thinner wire wrapped more revolutions is ideal. It would be awesome if this magnet had an led on it too. What shape should the magnet be?? I was thinking a cylinder which wouldn't be able to flip in the tube. I'm thinking it would have 3-4mm spacing between magnet and glass. :%

I did not posted this therefore I never mentioned any dimensions of anything. The author was. If he wants to make the magnet "float" it needs to be about the same diameter to avoid spinning. By the way I wasn't answering to your post. I think I know now why Grumpy is in your name. Maybe a similar word beginning with J will describe you be better.

Mike,

This member is "brainstorming" in another post regarding specific gravity. He is insistent if not self-indulging at the same time. I am not certain he will understand the magnetic force on a magnetic dipole depends not on magnetic induction, but the gradient of it, which is lacking in the center of the solenoid. Varying current is not going to vary this gradient so much. If you are close to the solenoid opening, then you do have this gradient. You can shoot something with a solenoid but the op wants some control.

It would be awesome if this magnet had an led on it too.

Then I guess the magnet will also need a battery added to it to power the led?

I've seen some videos of trying to hold a magnet in mid air and be position controlled and it's pretty hard to just hold the magnet in one steady position, but wanting to be able to change it's position vertically it gets very unstable and difficult if not impossible to control. The magnetic forces involved are not linear so the 'tuning' parameters for the PID control loop are virtually impossible to optimize for smooth control.

Not saying it can't be done on some level, but so far many have failed to demonstrate a perfected example. But perhaps you will be successful anyway as sometimes the most likely to succeed are those that don't know what can't be done.

Lefty

liudr:
Mike,

This member is "brainstorming" in another post regarding specific gravity. He is insistent if not self-indulging at the same time. I am not certain he will understand the magnetic force on a magnetic dipole depends not on magnetic induction, but the gradient of it, which is lacking in the center of the solenoid. Varying current is not going to vary this gradient so much. If you are close to the solenoid opening, then you do have this gradient. You can shoot something with a solenoid but the op wants some control.

Sorry to tell you; but you keep having it wrong...
The magnet is a real object with physical dimensions. As a cylinder it has height and volume. Its not a plane. That height makes it occupy a real space inside the solenoid and the top and bottom faces of it are not coincident with the solenoid center as will the ideal physical model you are describing. There is gradient from the top to the bottom face. There is also a force called gravity which will be pulling the magnet down and that will keep it off center if no energy is supplied to stabilize it there. On the other hand I never mentioned the center. That's you own creation; but that's Ok the magnet will stay at any position if enough energy is supplied.

Lefty,

Even pid would be difficult to change position. The force is not proportional to distance. If you just want to stabilize constant position then it is doable.

encryptor:
This does sound like a high energy project b/c I want the magnet to float.

I think you'll find that's quite hard to achieve, and if you're after that sort of thing I suggest you start off by leaving the glass out of the equation and go for a much smaller system.