View Full Version : Z axis assembly - unsupported rail for very limited movement

09-10-2012, 12:49 PM

I've posted in the FAQ section with a design of a grinder which uses a Z axis assembly taken straight from a CNC machine, here it is:


I've read a few threads about z-axis assemblies and quite often the main bit of advice is "use supported rail". Also swap the rail with the linear bearings, so that the rail stiffens the assembly. All great tips to make the machine better, although my gantry will only be doing 10mm of movement and most of the time will be sitting stationary as it's simply a support/adjustment element for the burr in the grinder, more than a moving, active part of a whole CNC machine. Keeping that in mind do you think the above is good enough? It's using 12mm rails, the plate with the linear bearings is 110mm wide and 160mm tall and everything will be made out of 10mm aluminium (12mm MDF for the prototype). As I want to keep the thing rather slim (various reasons), I'm planning to use a standard SS M10 threaded rod through an acetal block (50mm x 20mm x 20mm) and two 10mm Oilite bushings. I'm guessing that the main T plate together with the burr and other bits and bobs, minus the linear bearings, will weigh less than 1.5kg. I will need something to eliminate backlash so might end up with two slimmer 10mm lead screw blocks, one mounted stiff and the other with a bit of play and a spring between them to push them apart and then fix the second lead nut.

So main question: is it good enough or rubbish and needs re-designing?


09-10-2012, 11:15 PM
For such a small travel unsupported rails will be better than supported since the latter has a lower force rating in one direction (sort of).

If the Z-axis is sufficiently heavy, i.e. the downwards force on the axis due to it's weight (so about 1.5*9.8=14.7N) and any frictionaly forces, then you wont have any backlash since the force will not change direction - gravity is effectively providing preload. If this force isn't sufficient you could just add mass to the slide, although clearly this will cause the acetal nut would wear faster which may or may not be significant. A nut sprung like you suggest can work well, I did that once with brass nuts on M12 rod a very long time ago.

10-10-2012, 12:44 PM
Thanks for the reply Jonathan. Good to hear that the design isn't all bad:)

I re-did the calcs for the assembly and it turns out it's closer to 2500g, rather than 1500g (linear bearings are almost 1kg), which means more pull downwards and less backlash. I'm sticking with a single 20mm lead nut from delrin and will work on backlash if and when it happens.

Now the only thing I need to figure out is how to put the case together and make sure everything is straight and aligns. Using wood for the prototype might be more of a challenge than I though it would be.