Talk:Precision Machine Design

I see you only mention ballscrew so far. Is this for the torch table? Are you not considering gear and rack. It is the more common way to set up the long(x/y) axis on cnc tables. I believe it would be easier to machine from raw materials too.

Also I see you are looking into precision ground linear rails. This level of accuracy is overkill for an oxyfuel torch and for a regular plasma torch like your everlast. You can achieve sufficient accuracy with cheap ball bearing v-groove rollers or regular ball bearings on cold roll steel. Even the big companies like Esab have been building machines this way. It doesn't become necessary to use precision ground linear rails unless you get into high-definition plasma, which I believe the units start at about $25,000 for just the cutter. I would suggest designing a system that could easily fit both types of linear motion and the replicator can use whichever they choose.

My opinion in these matters is based on reading lots of other user's experiences (mostly on cnczone) and having put together a couple of tables myself, as well as observing commercially manufactured machine design. -LoadTest

For the CNC Mill/Lathe. Adding what I know and will cover rack and pinion later. Cheers!-Yoonseo

Oh, okay. Then ballscrews are probably the right choice if you can overlook the fact that they might be next to impossible to make accurately from scratch in a crude shop. Look into leadscrews vs ballscrews.

Considering ebay nor precision linear motion manufacturers(thk, skf, etc) might not exist or be accessible for every global village start up I would not use linear rails for the lathe/mill. Theres hundreds of thousands of mills and lathes with traditional box/dovetail ways that are producing high precision parts. Some of these machines are cnc from the factory, and some of them have been converted to cnc by diy. The surface grinder in the fablab doesn't use linear motion rails and they are often used for even more precise work then a standard mill. Once again, you might consider designing it for both, since alot of replicators with ebay accounts might opt for the linear rails since its easy to buy them and bolt them on as opposed to machining traditional ways perfectly parallel and smooth.

From what i've read, the number one failure in lathe/mill builds is the machine not being rigid enough. This results in surface flaws or worse in the part. Its a common cause of "chatter" as most machinist refer to it. Its also common to fill any cavities in the machine base with concrete or special types of polymers. -LoadTest

Yeah, exactly my thoughts. I may just have to go leadscrew, and to start off I definitely will go leadscrew- the modularity makes it easy to interchange later on anyway. Linear ways are similarly difficult like ballscrews and along your line of thought decided to go for an attempt at some small boxed ways- again, interchangeability is there, at least partially, so I'm not too concerned. I've recognized the major factors of precision machine design as precision interfacing, deflection, and vibration dampening, all encompassed by the words: precise rigidity. Oh also I haven't been looking into linear rails for the torch table (kind of curious what wikipage info led to that idea) but anyway thanks a bunch forthe info-Yoonseo