5

I need to move a rod with a 50cm stroke.

Almost no force is needed(the rod is just for display purposes). Speed is also not essential.

I was thinking of creating a very long DC solenoid but I am worried because I've never heard of such a solenoid. Is this impossible?

Avogado
  • 153
  • 1
  • 1
  • 3
  • Apparently we need some more informations about your system and the related constraints: do you already have the rod? What do you have, and what do you need? – clabacchio Mar 29 '12 at 08:41

7 Answers7

7

DC solenoids are hard to do at larger distances because the magnetic field decays with the square of distance. At 50 cm, you'd need a ridiculous amount of current. Such a solenoid is not impossible, but it is impractical.

The cheapest way to do linear motion (which, given your lack of speed and power constraints, is probably your primary motivating factor) is generally to use a stepper motor to move a belt or chain. 50 cm is actually quite a lot for linear motion, and this solution frees you from expensive long machined parts, like cylinder bores, pistons, threaded rods, and gear racks.

Something like this McMaster timing belt mated with a pair of these pulleys and with the (or whatever you're displaying) clamped to the middle of the belt should do the trick. Add a standard stepper motor, and be careful with your programming to avoid over-extending your belt.

Kevin Vermeer
  • 20,067
  • 8
  • 57
  • 102
  • 2
    It's worthwhile to note that some actuator assemblies are designed so they can be moved mechanically without the motor, while others are designed to hold their position when the motor isn't moving them. It's also worthwhile to note that surplus places sometimes have partial printer mechanisms available pretty cheap (for that matter, if this is a one-off, you could probably get an inkjet printer with a used-up ink cartridge for nothing and rip the guts out). – supercat Mar 29 '12 at 15:35
  • @supercat - Yep, WRT moving without the motor, a rule of thumb is that belts and racks can be moved, while threaded rods cannot. Printers are a great example of cheap linear motion systems; but I don't know of any that will go 50 cm unless you swap the belt for a larger one. – Kevin Vermeer Mar 29 '12 at 15:45
  • For weakly-linked flux, the field strength decays with the cube of the distance. – pericynthion Mar 14 '20 at 21:02
6

You could also use a circuit similar to a coil gun:

enter image description here

Without shooting to anyone (unless you want to :D) you can use separate coils to attract the rod from a longer distance without the need of a continuous coil.

I't just a matter of tweaking a little bit with the timing, but I've done it once, it's not so hard.

clabacchio
  • 13,501
  • 4
  • 42
  • 80
  • That's not a bad idea. I just have to be careful not to launch the rod out of the system :) – Avogado Mar 29 '12 at 07:15
  • You have two ways: the first is to put an obstacle at the end (obvious), the second is to use the second solenoid to "trap" the rod acting as a brake – clabacchio Mar 29 '12 at 07:17
6

Use threaded rod instead, spun by an electric motor.

tyblu
  • 8,207
  • 6
  • 41
  • 70
  • 1
    I think it might be good to note that these are commercially available: For instance, I've used Intelligent Actuator Robo-Cylinders before with great success, though they've probably got accuracy, lifetime, speed, and power ratings high above anything that the OP needs (and costs to match!) If the OP wanted to build their own, they'd want to use Acme threads, rather than standard threads. This sort of threaded rod is available at reasonable prices from McMaster-Carr: http://www.mcmaster.com/#general-purpose-acme-rods/=gv849d – Kevin Vermeer Mar 29 '12 at 03:06
5

An electric car antenna may be a ready-made solution

mikeselectricstuff
  • 61
  • 1
  • 1
    Yep; these are basically flexible rods wound on a spool connected to a stepper. Also, are you the same person as http://electronics.stackexchange.com/users/1681/mikeselectricstuff? – Kevin Vermeer Mar 29 '12 at 15:47
2

It may be simpler to make it pneumatic. A small pump, actuator, and electronically controlled valve will probably cost less than a solenoid of that size. As long as speed/force are "don't cares" you wont need a tank, and the pump will only need to run when you extend/retract the actuator. You could use a one-way actuator, but those usually require more pressure to overcome the spring, and you would still need a valve to let the gas exhaust.

Kris Bahnsen
  • 1,778
  • 12
  • 17
  • 1
    A simpler and more elegant solution but I fear it would be more noisy – Avogado Mar 29 '12 at 07:18
  • You could use a baffle on the exhaust path. However the pump would still be a good source of noise unless you could dampen in inside the end device, or find a quiet one. – Kris Bahnsen Mar 29 '12 at 16:12
2

If you have essentially no load requirement and you are looking for something cheap and easy to control, my suggestion would be an RC servo with a long horn coupled to a slot in the rod.

I don't have a way to draw it, but imagine that the bar you are moving is mounted in linear bearings that let it slide back and forth (bearings can be as simple as two pieces of nylon with holes drilled in them). Now a slot/hole is made in the rod large enough for the servo horn to pass through. Horn and bar are not attached in any way!

The horn is attached to the RC servo and the servo mounted so as the motor shaft rotates, the horn pushes/pulls on the rod. The necessary horn length you can calculate using plane geometry depending on how far you want the rod to move. There are more sophisticated ways of making the linkage (and reducing the backlash), but this is the simplest I can think of on short order.

Total cost should be less than a pneumatic air cylinder alone. I can get RC servos for $10 from the local hobby shop. Servo horn, metal rod and bearing material may add another $3. Actually, this is all stuff I have, so it's probably why I thought of doing it this way :-)

The servo itself can be controlled using one of the many, many examples online ranging from 555 PWM circuit to various microcontrollers.

lyndon
  • 4,386
  • 1
  • 17
  • 11
  • 1
    Remember that we're talking about 50 cm here! That would be a pretty hefty RC servo. – Kevin Vermeer Mar 29 '12 at 15:48
  • @KevinVermeer, nah, just use a continuous rotation servo and a rack and pinion style drive. You would need end sensors on either side of the rod's travel so it doesn't get overdriven. – Kris Bahnsen Mar 29 '12 at 16:09
  • 1
    @KrisBahnsen - That's different from an RC servo with a long horn! – Kevin Vermeer Mar 29 '12 at 16:11
  • Hah! It is, but I agree that the servo idea is not a bad one; it just needs to be tweaked a little bit to make it practical. – Kris Bahnsen Mar 29 '12 at 16:16
  • 1
    Ack! Yes, I somehow read 50 cm as 5cm. My bad. In that case I'd go with Kris Bahnsen's rack & pinion or a belt drive. – lyndon Mar 29 '12 at 18:22
0

You can use a dc motor (or stepper) with a sort of pulley on the shaft and 2 strings that a re attached to the pulley. When the motor turns clockwise one string tensions and the other loosens, if they are attached to something it will pull it to one side. And when the motor turns counterclockwise the opposite happens. That way you don't need a belt or coils or whatever. All you need is a motor. Whatever you do DON'T make a 50 cm long solenoid. That's a very bad idea .

Wissam Kadamani
  • 1