[musicmkr]

I'm new to CNCZone, and fairly new to the hobby. I should be finishing up my first desktop machine next week but I wanted to run a few details of my machine past you all and see what you think. Keep in mind as you answer that I'm only looking for accuracy of about .03.

Machine Basics:
-Aluminum 2"x4" tube frame.
-Moving surfaces 1" thick HDPE
-Part moves in Y Direction
-Spindle (dremel) moves in X and Z
-Machineable area 8"x7"x3.5"
-drive system is a belt drive, running 1:1 ratio to a 1/4"-20 rod in all directions.

Question 1 - The 1/4-20 rod spins in two mounted 1/4" needle bearings. (ie. I'm not turning down the shaft to remove the threads) Will this be a problem as long as load the bearings axially with a nut at either end?

Question 2 - Should I bother cutting a keyway into the plastic pully and threaded rod? Or can I simply thread in two machine screws through the pully hub?

Question 3 - I do not plan on using linear bearings. Instead I plan on having the 3/8" thompson rod slide directly through the 1" HDPE. Should this be a problem other than with long term wear and tear on the holes in the plastic?

Question 4 - Is it better to mount the lead screw A) directly between the slide rails, B) Opposite slide rail plane from part, C) Between slide rails and part?

Thanks for all your help!

[Chagrin]

1) With the size of your router and your accuracy requirements, you probably didn't even need any bearings. My router just has the 1/4-20 rod flopping around in slightly oversized holes and I don't see where it has caused any issue. Allowing the threaded rod to push (and move) the shaft of the stepper motor is probably the larger problem.

2) You might also try just jamming the pulley between two nuts (with lockwashers) before you start drilling or cutting at it.

3) Shouldn't be a problem, but try to keep a lot of distance between the two HDPE "bearings" that are sliding on the same rail to prevent binding. Even wooden bearings can be successful when done this way.

4) Directly between the slide rails. Again it's the issue of binding.