GearHeads Forum

Well, thats the problem. Actual specs are hard to come by, you normally see the spec for a belt instead, and the pulley rarely conforms
to the belt tooth. But, the GT5 for example has 7 arcs and the GT8 has 10. For a GT2 I interpolate from 22 points in a stored curve.
  I just checked and HTD's ARE just 3 arcs, a simple hole arc and two corner radii. So you could just drill circles and get away with it ,
Ill keep this onlist for a future GCode output for HTD's..

Art

Is your comment about several arcs with several centers true for both HTD and GT2?  The data I saw for HTD did not involve several arcs - just two plus the main circle.  That said, I am not sure it is correct.

Hi Bruce:

  The spec of those pulleys is of several arcs with several centers. A circle of holes would likely work, though may present
more belt wear that the real curves. Personally I think they make them that way on purpose..

  Ill keep in mind a way to simply derive them with drilled holes as a workaround..

Art

I was looking at the specs for an HTD 5m pulley.  A good part of the tooth profile could be generated with a 2.8mm drill.  I am thinking a very short drill (I have ordered some) and a delrin pulley.  An end mill slightly larger than the 2.8mm could be used to clean up the outer part of the tooth profile.  I think this would be faster than milling the whole profile with a smaller end mill.  So my suggestion is to add the capability to generate the drill pattern and the cleanup pass.  I have not seen the specs for a GT2-5mm pulley yet, but wonder if this approach would be helpful there as well. 

A note on the pulley specs - I used GT to generate a GT2-3 profile.  I exported a dxf and looked at it using Deltacad.  I was looking for arcs to dimension, but found none.  This was my attempt to reverse engineer the specs - no luck there.