Here are the pieces for the coil winder, partially assembled.
you can see the grain of the rapid prototype:
Already I can tell the bobbin is not well supported axially. I doubt it will be able to keep tension with 12 gauge wire. But we can test thinner wire on this iteration.
Putting it together:
As expected, the axel needed reinforcement:
Doing some quick checks. The stepper motors are supposed to take exactly 400 steps to complete a revolution. However, when I program the bobbin to take 400 steps forward then 400 steps backward, it appears to come just short of a full revolution! WTF! It looks like it’s closer to 415 steps per revolution. But I can’t trust that number to be accurate over many revolutions.
I’m learning RAD, a gem for controlling the arduino from ruby. Very cool.
Prometheus Fusion Perfection 
Your set up does not have enough torque. You are skipping steps.
May I suggest a switching set up which increases motor current to 141% on the half steps?
Or a bigger motor with 2X the torque (safety factor).
Please contact me directly (e-mail on my sidebar) when you have stepper motor questions. I have had some experience. Also any electronics questions.
Yeah, I’m redesigning the bobbin. I plan to have one side of the bobbin be a large gear, then use a timing belt for gear reduction.
This is why I’m building the reprap, that bobbin was $370, all the parts together were $1K. Way too much for the amount I’ll need to iterate. If I can get the reprap working well (which seems to be possible), I can make all sorts of specialty tooling like bobbins and jigs.