We're started building our first router. This design is made from aluminium plate. We don't have access to a milling machine so one of the criteria was to be have simple parts.

Progress to date:

http://www.homecnc.org/images/thumb/4/4b/2006-08-28-IMG_0930.JPG/1024px-2006-08-28-IMG_0930.JPG

This is the partially assembled machine (a number of parts still need boring work).

Rhino design and progress is documented at

http://www.homecnc.org/index.php/CNC_Milling_machine


The end result should wind up looking like...

http://www.homecnc.org/images/9/91/CNCMill-v0-2-p1.jpg

i shouldn't write about stability, since i didn't finished my design (okay - design finished, but not became real machine yet), but...
i would suggest you increasing bearings spread in all axis, adn adding additional plates on every bearing - one is not enought (in my opinion at least)
you should consider making z axis with one drive - two, if missed steps occour may lead you to a dead stall
if you want to leave 2 threaded rods on z axis, then at least use one engine and geared belt to transfer power
if stepper will miss a step - both rods will

Agreed, the bearing spread is a concern, but it's easy enough to change it later so we'll start with a 100mm spread and see how we go. The linear bearings are reasonably tight and we're intending to pre-tension the rods so we'll see how we go. Obviously increasing the spread will reduce the travel.

Re the two steppers. If this turns out to be a serious issue I'll add encoders to the Z shafts to detect and correct this. The thread pitch is fairly low so a cheap encoder that only does 16 or 32 pulses per turn would be adequate.

scared of geared belts are we? you're absolutly right - they have so many tooths that they have to bite!
2 steppers and two encoders + electronics controlling them will be cheaper than 3 gears, one stepper and synchro belt?
wierd logic behind this!
oh - i failed to mention 2 couplings versus 0 in belt approach

scared of geared belts are we? you're absolutly right - they have so many tooths that they have to bite!
2 steppers and two encoders + electronics controlling them will be cheaper than 3 gears, one stepper and synchro belt?
wierd logic behind this!
oh - i failed to mention 2 couplings versus 0 in belt approach

Yes, I'm deathly afraid of geared belts. In part because of a terrible childhood accident, but also because I don't have any.

Whereas, I do already have the electronics to run 4 steppers, 2 encoders and stepper motors to suit. :)

So for me, moving to a geared belt setup would be a significant increase in complexity. I may eventually go in that direction, but not yet...

point taken:] hope you'll finish your design soon