[Swede]

Thought you IC types might get a kick out of this... it is one of the smaller CNC machines you'll ever see. It is a CNC ignition coil winding machine, powered by 2 stepper motors salvaged from 5.25" floppy drives. The mission was to wind 30,000+ turns of an incredibly fine-guaged copper wire in multiple layers on a Delrin spool. This wire was abysmally fragile, and the slightest drag caused breakage. The wire had to be laid down via 2 shafts in a CNC gear arrangement, with the pitch of the laydown feeder being the diameter of the fine wire - in other words, as the bobbin rotated once, the tiny tygon tube which fed the wire, traversed in X a distance equal to the wire diameter. Thus the coil was laid down in an appearance similar to a spring with no gaps.

Each layer of winds was insulated from the next with a section of kapton tape, very tedious. This was needed because the outer turns of wire ehibit a huge potential (>10,000V) relative to the first turns, and the spark would arc internally without the insulating layers.

The stepper drivers are 1A Ramsey stepper motor kit boards. The software I wrote using VB, directly "printing" to the parallel port.

[Swede]

Here is a picture of the laydown assembly. The wire feeds from the back, and over several very small plastic pulleys, ultimately feeds into a tygon tube and then onto the bobbin. The carriage is on a 40 TPI shaft, and can be repositioned as the coil "fattens". The exit point of the laydown carriage needed to feed the wire directly onto the bobbin.

[Swede]

I've only made 3 or 4 coils, and even with CNC they are very labor intensive, mainly due to the kapton tape insulation. I'd like to find a thin enamel which I can brush onto each layer to insulate, which would wick into the wire and replace the tape.

One of the coils was mounted into a gas turbine "buzz-box" ignitor. Pressing a button on the box causes a continuous buzzing spark through a 1/4 X 32 plug. This caused instant ignition in my MW54 gas turbine, but the stray votages played hell with ECU electronics, so it was shelved in favor of a conventional glo-plug.

[bb99]

Nice Work!

What is the overall size of the machine?

--bb99

[HuFlungDung]

Very interesting, Swede

The professional motor rewinders which we do some work for, use "magnet wire" that is pre-insulated with some kind of high quality lacquer insulation. Is this what you are using, but adding seperate insulation too?

[Swede]

bb99: Thank you! The uprights on the machine which form the supports for both axes (bobbin and laydown carriage) are 2.5" apart. Those little motors are gutsy for their size. I remember (this was about 1999) going through a big box of 5.25" drives at a PC store and buying about 5 for maybe $3 each. There were a number of different steppers in them, some good for this type of project, some not so good. These are 1.8 degree unipolar steppers. Some of the disk drives had steppers which were integrated with their leadscrew device and weren't suitable.

HuFlung, the wire is enamel coated, 44 guage, 0.0026" dia. It is insulated, but the trick with the coils is that each LAYER of winds must be insulated from the others. The coils generate about 20 kV, with a steadily increasing potential from one end to the other. The thing to avoid was allowing the last few turns to arc between layers to the first few turns, as 20,000V will jump a big gap, and once even one arc occurs inside, it leaves behind a trail of carbon, which conducts, making subsequent arcs automatic.

Here is a source of magnet wire online:

http://www.planetengineers.com/defau...Wire%2C+Magnet

Get this, the 54 guage copper is so fine that 1 pound will unwind to a length of 156 miles!

If anyone else is interested, get Bob Shores's wonderful book "Ignition Coils and Magnetos in Miniature. Bob is an older, ultra talented gentleman who has plans for a number of sweet little engines, which I noticed a few of the guys are working on. His book is worth every penny, as it has great info not only on coils, but on ignition circuits and methods for small engines.

Here is one last pic, showing a finished coil potted into a delrin tube. These coils can also be mounted in a wax filled box (like a model T coil) for a cool scale appearance next to a small farm or hit and miss engine.

If I can find some thin liquid insulating enamel, I'd like to resurrect this device and crank out some more. Maybe there's a market for them, as the coils you can rip off from a wrecked 2-stroke garden tool almost always have electronics potted inside, and they aren't suitable for a straight CDI or kettering ignition for our small engines.

[anoel]

Very cool....

I'm into winding guitar pickups. Not exactly the same thing but It involves wrapping 9,000 turns or so of 42-44awg wire onto a bobbin.

I made my own winder. (Cut most of the parts on my homemade CNC machine.) The machine uses a cam driven traverse. It works very well. I've got to clean up a few things but the pickups I have wound sound and function excellent.

[bb99]

anoel,

What is the LED number on the right? Number of windings?

--bb99

[anoel]

It's a digital counter, it can be used as an up or down counter. I use it as a down counter. I placed a small magnet in the faceplate and use a halleffect switch to trigger the counter. Once the counter reaches zero it triggers a relay to shut off power to the motor.

[Bloy2004]

Cool! I like that machine too as well as Swede's...fascinating!

[anoel]

Swede, For pickup coils we pot them in a mixture of 80% Paraffin and 20% BeesWax. Heated up to about 140-150 degrees F. Some of the guys use a vaccum pot to sink th ewax deep into the coil. But just soaking the coil in the wax does a pretty good job on it's own since the wax is of a water consistincy at that temperature. That would work on your coils unless the temperature gets above the wax's melting point.

If it does, you might be able to score some Formvar varnish. A simple vaccum can be made to pot the coil deeply with varnish. THe only draw back is you've got to bake the coil for a good while to cure the varnish. The wax will dry on it's own.

[Swede]

Thanks Nathan. Ignition coils can get pretty hot, so I'd like to try something pretty resistant. What exactly is Formvar? The vacuum is no problem as I have a chamber made from surplus parts, ditto the oven, I've got a nice one for tempering metal. The real trick with HV coils is the interlayer insulation. I think the best would be a two-fold method, with either the kaption tape or some polymer between layers, followed then by the potting. In this size with the ultra-fine secondary, and no potting, even ignoring the arcing, the vibrations would ultimately cause failure. That wire is so darned fine!

Also I never noticed earlier but the prices on those enamel wires at http://www.planetengineers.com/defa...=Wire%2C+Magnet REALLY skyrocket above about 43 guage. THOUSANDS of $$ for a couple pounds of copper! WHOOOOOOO! But one spool of #44 or #45 will last a lifetime, darned near.

I'm psyched to make more coils now. I need to drag out my source code and refine the software.

Nathan, I like your rig and your concept for the hall IC/counter cutoff device. My problem is I have to halt between layers to do the insulating thing, so my software has a built in stop at the end of each layer. It'd be great to totally automate the whole thing, maybe that's be a fun goal for the future.

[amoretti]

Swede,

I'd like to send you a PM but your message box is overfull. I am very interested in your CNC winder. Please PM me back!! Cheers, John

[nemesis]

... coils generate about 20 kV ...

Wind a bit more and you could make yourself a taser.
They are basically CDI systems with exposed spark gaps.
75kV would be effective I should think.

I had a go, but like you say the internal breakdown is hard to fight and once it flashes over it is permanently wrecked. Definitely recommend the mylar tape on every layer, followed by potting in paraffin method

[LoginIsFutile]

I want to build the cnc coil winder machine, please sen to me the details to the construction and the software for the wind to [email protected]

[signIT]

Hi,

If you own a cnc router and need to wind coils for any purpose, maybe a exsisting cnc router can be useful.

The wire spread, wire tension and wire holder can be made as a option part that can be mounted on the z-axis mechanism.

The x or y axis driver can be connected to a stepper motor with the coil or bobin holder, this part can be attached on the x/y cutting table.

But, how to program this application to run as a coil winder, it would be very nice if anybody have tried this.

I have read a lot about making a coil winder but it seems that the most of them are standalone machines, but if you already own a cnc router it would be possible to convert it to a real nice transformer/coil winding machine.

[lerman]

Consider the area to be filled a rectangle, with the center of the bobbin at the bottom and the ends of the bobbin at the left and right.

Then lay down two turns starting at the left edge. The wind backwards diagonally to lay down another turn nestled between those two (but higher). Now go diagonally back and forth laying down turns in a triangular pattern until you have the requisite coil diameter.

Then continue diagonally back and forth increasing the length of the coil until you are at the end. Finish up by filling int the triangle on the right.

This pattern will have the characteristics that the voltage between turns that are near another is small. The two ends at the maximum voltage difference will be far apart.

Does my description make sense? Will this work?

Ken

[Ken Stuempge]

Hi all. I have completed the mechanical part of my coil winder, using most of the concepts Swede posted at the start to this thread. Here are some photos of the winder . . .The steppers are 200 step/rev, the feed rod is 1/4-20. Max length of the coil to be wound will be 3 to 4 inches. Most of what I plan on winding will be 1" long, for scale IC engines.

[Ken Stuempge]

Here are some detail pix of the feed portion. I made the pulleys out of Nylon, the small white block shown on the top right of the two left photos is UHMW, which will have a small notch from an exacto knife to serve as a guide for the #44 wire instead of a hypodermic needle. Hopefully that should work as well as a hypodermic, with no risk of scoring the insulation, and provide a very precise guide for the small wire. The small white block on the bottom is the feed block that serves as an eyelet on the front, and is also UHMW, with a #60 drill hole.

[Ken Stuempge]

Finally I have a pix of the take-up rod for the Bobbin, and how I plan to provide tension to the wire. The feed spool is mounted on ball bearings, and the coil winder itself will just sit on top of the aluminum plate, so the distance from the feed to the winder can be easily adjusted. If needed, I can add some hi-tech wire tensioner to the plate as well, if my friction adjust does not work. The tension can be adjusted using the small brass thumbscrew to move the spring "finger" closer to the spool rim.

The rod for the bobbin has two springs and flanges to apply pressure , and allow the bobbin to be located anywhere along the length of the rod.

I plan on using a PIC 16F870, with a 16 key keypad and LCD display. The display will prompt for wire size, length of the coil to be wound, and total number of turns. The two steppers will work in tandem, to provide linear movement / bobbin revolutions to space the wire properly. Winding speed will be controlled with a pot to feed an one of the ADC inputs on the PIC. The process will be to wind one layer, pause so insulating tape can be applied, then a key stroke will restart, moving the feed in the opposite direction, repeat, etc .

I have not started the software yet, but will today. Hopefully it won't take too many months to completehttp://www.cnczone.com/forums/images/icons/icon11.gif

I'm sure I will encounter more than enough problems , and will probably ask for help.

Ken