Building a CNC coil winder

Thread Starter

electronice123

Joined Oct 10, 2008
339
Hi everyone,

I am working on a project to build a cnc coil winder just like the one in the link below, but more accurate.

http://www.theprojectasylum.com/electronicsprojects/automaticcoilwinder/automaticcoilwinder.html

I will be winding coils with 36AWG wire.

The website project used a 7.5 degree stepper motor with a 32TPI leadscrew, so the travel per step is .00065". From my calcs using 36 AWG wire the wire would overlap every 10-15 turns. NOT GOOD.

I have not used stepper motors before so this is all new to me.

MY QUESTION:

Are there any circuits or drivers I could use in place of or add to the driver shown on the website to utilize micro-stepping?
 

THE_RB

Joined Feb 11, 2008
5,438
What's the problem with 10 or 15 turns?

There is absolutely tons of resolution available on the wire positioning motor, even without microstepping.
 

Thread Starter

electronice123

Joined Oct 10, 2008
339
Maybe I'm doing my math wrong. 36awg wire with heavy insulation is .006" dia. So if the accuracy is .00065" then wouldn't every 9th turn result in an overlap? .00065/.006=9.2 turns?

Also, I purchased the magazine article. They had a coil wound with 32awg wire that looked horrible. Lots of overlaps.
 

kubeek

Joined Sep 20, 2005
5,794
First of all, doesn´t the wire basically position itself when winding if you have correct tension in the wire and a tiny bit of lag in the guide?
Anyway, your assumtion seems wrong to me. That accuracy will not add up, i.e it doesn´t mean that after ten turns you will be 6.5mil off and after 20 tuns you will be 13mil off, it means you will be +/-0.65mil off at any time, but all the way from start to end about the same. And since this is about 1/10th of the wire thicknes you should be ok.
 

atferrari

Joined Jan 6, 2004
4,764
Yes. The PCB milling machine I was planning to buy, started to be offered with software developed for USB just less than two years ago.
 

shortbus

Joined Sep 30, 2009
10,045
@electronice123, you might want to consider using one of the 'linear stepper motors' for the wire guiding drive. They have the threaded part built in to them, and are much more accurate than a screw thread. http://www.haydonkerk.com/?TabId=66 There are many companies making making them, the link is just one of them. I got mine on Ebay.
 

kubeek

Joined Sep 20, 2005
5,794
From what I recollect from using a parallel port in windows to control motors, it is a bad idea. For anything faster then a few steps per second it is useless, as windows often gives focus to other tasks and halts the execution of your control program, which results in missed steps.

I made my CNC mill with ATmega controller which is fed over USB to serial converter, this makes the operation much more predictable.
 

Thread Starter

electronice123

Joined Oct 10, 2008
339
Thanks everyone for the comments so far.

So, does anyone here any suggestions on how I could build a better cnc coil winder?

What should I do instead of the project on the website?
 

kubeek

Joined Sep 20, 2005
5,794
If you don´t require the turns to be spaced apart, why not try using the stepper just for the spindle and guiding the wire manually? I think the mechanics of the setup is the most important part, so I would start with making a few coils manually like that and decide what needs to be done to make it work right.
For example this vid looks nice http://www.youtube.com/watch?feature=player_embedded&v=uX2iCRdSFmI
 
Last edited:

atferrari

Joined Jan 6, 2004
4,764
If you don´t require the turns to be spaced apart, why not try using the stepper just for the spindle and guiding the wire manually? http://www.youtube.com/watch?feature=player_embedded&v=uX2iCRdSFmI
Stubborn as I am sometimes, I refused to accept advise of NOT using manually guiding the wire.

After completing my machine, after three painful hours of testing the most varied things to guide the wire manually I had to accept that the suggestion was correct and my work (a lot) useless.

Please don't.
 

mcgyvr

Joined Oct 15, 2009
5,394
There is really nothing wrong with what you started with..
and
Timing issues with windows should not be a problem at all for this application.
and
As long as the distance per step is less than the wire diameter you are good to go..
 

Thread Starter

electronice123

Joined Oct 10, 2008
339
Mcgyvr,

I will go ahead then and build the cnc winder.

I have one more question though.

If I buy a usb - db25 cable would it work?
There are some on eBay that come with driver cd's-they state they will convert any parallel printer to usb.
 
Last edited:

mcgyvr

Joined Oct 15, 2009
5,394
If I buy a usb - db25 cable would it work?
There are some on eBay that come with driver cd's-they state they will convert any parallel printer to usb.
Probably NO..
Not positive but I believe that goes the "other" way.. If you have a USB printer and somehow no USB port on your computer then thats what you want..
 

Rbeckett

Joined Sep 3, 2010
208
Why not just get a small mini lathe like a 7X10 from harbor freight or EBay and make a lathe winder and use the feed to wind the armature. You can get them on ebay fairly cheap and some even have a tach, so you could add a counter so you get your winding ratios correct. No need to build one if you dont have to. Otherwise all of the boxes, boards, motors, drivers, fixtures, jigs, castings and stuff will probably cost more than the lathe and be a system that no one besides yourself can get to operate.

Bob.
 

THE_RB

Joined Feb 11, 2008
5,438
Maybe I'm doing my math wrong. 36awg wire with heavy insulation is .006" dia. So if the accuracy is .00065" then wouldn't every 9th turn result in an overlap? .00065/.006=9.2 turns?
...
Yeah I think you misunderstood a little bit.

The positioning motor needs to move an *average* of 0.006" per every turn of the coil.

The positioning motor moves in steps much smaller than that, ie steps of 0.00065".

The controlling software does math and always picks the closest positioning step to the perfect desired position. So there is no accumulated error, error at any time will be less than half a positioning step, and the average positioning speed will be perfect over time.

As for wire overlaps, those are most likely the result of setting the positioning to the wrong speed, or of bad mechanicals, specifically wire tensioning.

For good results you need to have really good wire tensioning, that you can adjust finely, and will be reliable at exactly that tension. If the tension varies up and down you will get some ugly results. :)
 

Thread Starter

electronice123

Joined Oct 10, 2008
339
The RB,

Thanks for the info. The coil tensioner is one thing I am still trying to figure out.

Any ideas on how I could make a wire tensioner or where I could buy one for cheap?
 
Top