Ugliest, cheapest auto traversing coil winder ever - mine!

[peskywinnets]

Ok, so I decided to make an auto-traversing coil winder on the cheap. I ripped open an old dot matrix rrinter for a 7.5 deg stepper & it went from there.

Essentially, the main winding motor is an old DC cooling fan. I've hot glued a drill chuck into the centre of it (tricky!). I'm feeding magnet wire onto the main motor via a nylon threaded (helix) rod...the rod is turned to & fro by a stepper motor. The stepper gets steps from the main motor because I mounted magnets on the spinning bit ...as these magnets pass a reed switch on the fan chassis, you get a switched effect to a PIC chip. I then learned how to program PICs. (this is a pretty good route because no matter how fast the main motor spins, the nylon threaded rod stays in synch ..due to it deriving it's pulses from the main motor magnets)

All in all it's costs me about $10 (for the drill chuck, a nylon rod & PIC).

I gave it its first trial run this morning - it works pretty well, except when changing direction! (perhpas there's a market for very very long coils with only one layer?)

What happens at the end of the coil, the nylon rod is reversed by the stepper & starts turning the other way taking the feed wire with it...however the wire feeding onto the main motor has some reluctance to the change of direction..though it eventually yield when the wire has moved so far away at such an angle to force the change. This resulted in an uneven first layer & subsequent layers.

(this isn't mine but it's pretty much the same concept - http://www.youtube.com/watch?v=Q_Cd_5zTt9w )

I'm sure this is a conventional problem that everyone hits with such a device...but what's the solution?!!

(I'll try & get a youtube video up soon to show the problem - remember though it's butt ugly & wont win any Engineering awards (it might just win the "cheapest coil winder ever award" though!)

[David Schwab]

That's a real interesting way to do it!

The thing about the traverse is you can't have linear motion across the bobbin or the wire will pile up on the ends. You have to spend more time in the middle and less time at the ends.

With mechanical traverses they use a heart shaped cam.

[peskywinnets]

Thanks for the reply....not sure that approach will work in this instance (eg presently I have more copper in the middle...therefore I need my program to spend less tim in the middle & more at the ends!)

ok, I've since applied a bit more thought to this & will adopt two approaches...

1. Use a bobbin...this will force the wire up one layer as it gets to coil side extremeties (on account it has nowhere else to go!), before traversing the other way. I had been simply winding onto a pencil - as a test run - the lack of a Bobbin is troublesome!

2. I will modify my PIC program to dwell a little at the extremeties. Essentially, the PIC program knows when a wire feed traversal extremity is reached, but presently, it starts traversing the other way as soon as it hits that point. I reckon I need my traverser motor feeding the copper wire to wait one revolution of the main motor. Again, this will allow the wire up one layer (on top of itself) before taking off in the other direction)

Not sure, if I'll get time to update the program for a day or two. I'll try & get a video up of my 'solution' soon (but you've all got to promise not to laugh....no, I mean *all* of you)

[Bruce Johnson]

That's a real interesting way to do it!

The thing about the traverse is you can't have linear motion across the bobbin or the wire will pile up on the ends. You have to spend more time in the middle and less time at the ends.

With mechanical traverses they use a heart shaped cam.

David, I'd disagree with that. My winder is a true linear traverse and, if anything, it will wind a little too heavy in the middle. That's mostly because I tend to set the stops a bit shy of the ends for safety, to minimize the chances of it snagging and throwing a loose loop. But if I run the stops to the ends, it will lay it down quite evenly by itself. And some of my coils are 14,000 turns, which ends up more than 1/4" of wire on each side.

The reason why most winders use a shaped cam is to compensate for the geometry of a pivoting arm, and the take-up of slack as the linkage changes direction at each end.

[David Schwab]

Bruce, I have no experience with auto traverse mechanisms, so I'm just repeating what I've read.

You have a very interesting setup. Your traverse moves from side to side and isn't on a pivoting arm, so that makes sense.

When I get a chance I want to come up with some kind of traverse mechanism, as well as a tensioner.

Here's a winder a friend of mine has which he's fixing up. It also has a side to side traverse, and a bunch of different shaped cams.

[JGundry]

I have one of those old Coweco winders. It almost identical to the one in the photo but without the tailstock. The cam on mine is sloped and you can do from a little under 1/4" traverse to a little over 1/2" all with the one cam. It is a cool little machine. Mine was made in 1959.