Welcome to the SeeMeCNC Community

I needed a Bowden setup because my custom linear carriage had my extruder too low and eating the belt. Please ignore the duct tape. I posted two threads before but got no replies, after I realized it was top secret information I built my own. Clicking on the pics will make them better. This is what I came up with:


[img]http://s8.postimage.org/kw8m8voqp/picture023.jpg[/img]

[img]http://s8.postimage.org/yrbi538jl/picture024.jpg[/img]

[img]http://s8.postimage.org/nt08mwjy9/picture025.jpg[/img]

[img]http://s8.postimage.org/u7z9jkqo1/picture026.jpg[/img]

Why is the belt twisted in the back?

Looks good, but you really want the side of the belt that's pulling the carriage to be as close to horizontal as possible, any deviation from that will make motion none linear and you have increasing dimensional errors as you move away from the center.

That's jus how I mounted it, There were already holes there so I took the easy way . Doesn't make any difference that I can see .

tom10122,

The effect of that nonlinearity will be impossible to see but maybe not impossible measurable. As an example, using Pythagorus theorem A^2 + B^2 = C^2, lets assume that the belt rises 10mm over a 200mm length (I estimated this from your photos). So C (hypotenuse, the length of the belt) = 200.25mm. This means that if you instruct the X axis to move 200 mm, it will actually only move 199.75 mm. That is .25mm over about the full range of the X travel.

Note, this is a really simplified calculation, the situation is actually worse since as you get closer to either end, the "triangle" formed by the belt and its offset from parallel is worse so the error is multiplied at the ends of travel.

cheers,
Michael