Intro to Open Emag
After spending roughly 15$ USD on ingredients and printing this design (with 3$ of PLA plastic), you can own a high quality coil wrapping machine that would have costed many times more in the proprietary world. We all deserve good electromagnets. My reward (eventually) will be to see the photos of other people's successful renditions.
WARNING: This project is JUST THE MECHANICAL apparatus. I do not supply code for controlling the device because there's too many ways to do it. I discuss many options for controlling the servo and motor, but I do not hold your hand and walk you through to finishing the project. Most people reading this are interested in doing something really nerdy and specific, so this modular approach is likely a feature, not a flaw.
This is an OpenScad project that uses a central file (openEmag.scad) that calls upon all of the other files to accumulate a ready-to-print kit that can be assembled (along with a few products) into an electromagnet-wrapping machine. openEmag.stl is the entire kit, rendered such that no support is needed, and the model will fit on printers like the CraftBot (bigger bed than usual). If your bed is too small, you'll have to print the individual parts and most of these parts, I only provide the Scad file. OpenScad files can be converted to STL easily and this allows anyone to edit any part of the design. Most users of this project would like to change the diameter of the wire spool (spoolform.scad) which sets the inner diameter of the electromagnet. Thanks to OpenScad, I don't need a Customizer. Just edit my well-written code! Enjoy the demo video on this page.
Open Emag Controller Project
Something must power the motor while telling the servo to go back and forth slowly. In all cases, the predictable turning rate of the motor means that time can be used as a measure of turns: (time to wrap)=(number of turns)/(rpm) in minutes. There are many ways to operate this machine but I can think of 3 main categories:
- Human controller: You could skip the purchase of a servo and just use your hand, similar to what I do in the demonstration. This would still be much faster and better than manually wrapping without the device. You could even save plastic and comment out the servo mount from the design. Not to hurt your feelings, but this method will yield lower quality results and far less coil consistency than the next two methods.
Analogue controller: One could make a circuit that produces the control signal for the servo, just making it slowly wave back and forth. There are these things called 'servo testers' which do just this. Along with a switch and a pocket stop watch, fairly accurate electromagnets can be created.
- Digital controller: You can use a computer, like the Raspberry Pi or the Adruino . There are MANY advantages to using a micro-controller, like a variety of wrapping patterns and automatic shut-off when the number of turns has been reached.
Separating the controller project from the mechanical design allows people to choose how much they want to spend and what features they want.
Electromagnets are important in modern life. The average person in modern society relies on electromagnets daily. They have been out for long over a century, and electromagnets are key to all electric motors, all speakers, and all magnetically locking doors, for a small set of examples. There probably are many more inventions that could be created to benefit the world using electromagnets but the public is generally unaware of them. If more people know about this century old idea, people will understand much more about what's going on around them and people could even make new things. It is very likely that the plethora of conveniences from electromagnets is not even close to being realized yet. Electromagnets are the only magnets that can turn off or reverse themselves or change strength on demand. I get excited just thinking about that !!
Why use the machine?
Good electromagnets have hundreds, or thousands of turns. We want to push the limits. No one (in their right mind) wants to count hundreds or thousands of anything. Also, this offers repeatability. You can make 10 electromagnets with 600 turns each, and it would ideally take a little over 30 minutes with this machine (at 200 rpm), that's 3 minutes per electromagnet. Doing that same task by hand?-- well, that's the alternative.
Audio explanation: The phone was sitting directly on the table. The wire spool is bouncing around on the spindle (far left) and that sound is traveling directly though the table to the phone.
Note that you can find each of these products from different competitors, etc.
One screw will be the axle for the 'Cursor' part which directs the wire onto the spool. A few more are used to hold the coil motor in place. Get the accompanying m2.5 nylon bolts. m2.5 is the keyword for the right diameter and you will need 25mm length screws for the coil motor. For the other parts, you will want to snip the nylon screws shorter with strong scissors or a snipping tool like I use in the photos above.
Turns the spool, and needs a strengthening gear ratio to work against the friction of the rubber band. Here are 3 options:
There's lots of gear motors like this, but I can't find one online as fast as the one I own! I got mine from inside a Microcenter, but it's not on the Microcenter website. I'll report the speed of the robotdigg option when I get it. Generally, you will want 60 rpm to 200 rpm because slower than 1 turn per second (60rpm) will be on the slow side, and faster than 3 turns per second is getting toward the weak zone. Some of these motors take more than 6v input, and the rpm will increase linearly with voltage. So if you double the voltage, you will roughly double the rpm. Warning: high power and high stress can shorten the lifespan of a gear motor too so be careful with your need for speed!
Moves the cursor through a simple lever system. This CAN be skipped if you want to move the cursor with your hand as I do in the video.
30 AWG is a great (VERY thin) diameter for electromagnets for 1v to 9v experiments.
22 AWG (which is thicker than 30AWG) is getting close to the limit of wire thickness for this machine. This thicker wire will handle higher voltage (without melting the coating on the wire) than 30 AWG. You can actually pull the whole coil off when it's wrapped, as opposed to the 30 AWG wire.
There is more info on the effects of wire thickness in the FAQ section.
3 inches of 14 gauge wire and Rubber band
You will finally need a hanger wire or similar thick wire, which is that green wire in the photos. If you really need to buy that, it would be 14 gauge wire, thick enough to connect the servo arm to the 'cursor' part of this project. The design has two hooks for connecting a rubber band which will provide tension for the wire as it's wound onto the spool.
Follow the photos in the slider above. There are no words, but the images are showing each step of assembly, in the correct order.
Firstly, I do not know your plan with the finished electromagnet. You might want to leave the coiled wire on the spool form , or you might plan to slide the coil off... You likely will want to modify spoolform.scad to set the size and width of the spool form, and you can easily add another wall on the other side of the spool form.
If you want to use thicker wire than 22 AWG for this machine, you may need to print with a material that is stronger than PLA and use a stronger rubber band. I have not tried this, and I don't plan to try it myself. On the other hand, you can use arbitrarily thin wire easily (must be insulated).
When you put the wire through the guide (intersection of rubber band and magnet wire), a good trick is to make a hook shape with the end of the wire first. Look carefully at the photo where I'm about to put the wire in, it's hooked. THEN put the rubber band on. The guide is aimed upward toward the spool form to minimize the shape memory problem. It reduced the shape memory problem (for thick wire).
- You can get an accurate number of wraps in many ways. One way is to use an external device that can count the number of turns for you. You cannot just trust the rpm label on the motor mostly because of the rubber band which helps keep the coil tight. So the real rpm will be slightly less than the specification. If you aren't using an additional device to count the turns, you can just measure the rpm carefully in a test run (record a video, play in slow motion, count, etc. do you want accuracy or not ?). If you don't care (at all?) about the number of turns, you can still make near-exact copies of electromagnets if you just measure by time only.
Feb 29 2016: made all of the holes wider so it's easier to slide screws into place. I also made the spindle holders simpler in design to avoid weak spots which were happening with my Craftbot printer. This updated version is what you're seeing in the photos in the slider above.
March 18 2016: I cleaned up the spoolform.scad file so that it is MUCH easier to change the properties. Now, You can set the radius and the depth of the cylinder, and the comments tell you the operable ranges. The spool form also has much better starter holes (5 of them) which is in the back plate instead of going through the core. I also renamed coilMaker.scad to openEmag.scad and I renamed coilMaker.stl to openEmag.stl.
- March 20 2016: New cursor design (updated cursor.scad). Diamond shaped cursor holes increase the accuracy of the wrapping pattern. This has been rendered into the OpenEmag.stl file.