though my models aren't as nice as MrBean's, I hope they convey
the idea.

plasma table, 4'x4'. planning to use 20mm rounds for table traverse (X?), one on each side, with linear bearings.

two 15mm rounds will carry the carriage across (Y?).

i want to keep the gantry as light as possible. i'm hoping light enough so the 'rails' (round stock) will act as both structural members and linear slides.

the belt is fixed to the carriage, between the rails. i plan to add a 'tab' to the back side of the belt and use that to activate limit and home switches located on the gantry supports, one on each side.

the gantry support (if i'm using the right term) is an inverted T shape and will hold two linear bearings each, about 20cm apart. hope thats enough to avoid racking.

the X axis will be driven exactly the same way.

no driven Z axis for now. just a torch holder on a dovetail slide.

500oz steppers attached to 14T pulley, belted to 28T pulley (ie 2:1) which drives the idlers (one on each end of each axis) that are also 14T. these sizes were on hand.

my numbers show a poor-ish resolution of 0.8 mm per step (200 steps/rev).
200steps/rev x 14T/rev x 12mm/T= ~ 1.2 steps/mm or 0.84 mm/step.

considering my plasma cuts an 1/8" kerf, i'm wondering if 0.8mm is ok for now.

this is my first CNC project. if all works well i can change the pulleys for more resolution?

what do you think?
-anthony

caught an error in my math?
there is a 2:1 ration between the small and large pulleys,
so 0.84mm/step is actually 0.42mm/step?
-anthony

I'm only guessing but you might get sag using those 20mm and 15 mm rounds when traversing midrange....depending on the weight factor. Your nozzle may be too close midway or too far away at the travels ends. I would consider supported rails(rounds) and open bearings. ....there goes the budget!

The steps should not be an issue. My geckos have 2000 steps per rev. There is no need for you to use a 200 stepps per rev system. Most boards have selects for micro stepping.