7 Axis Swiss Mill-Turn (DIY)

[Krispee]

I think lead screws are a little more tolerant of chips getting in them because they're self-wiping. The nut slides along the screw and therefore pushes most large debris off. Ball screws usually have some sort of wiper on them, but it's usually fairly soft and probably wears out eventually, but the balls roll over the screw, so if there's debris that makes it past the wiper, the balls may roll right over it.
Plus lead screws require more frequent lubrication, and putting oil in the nut will tend to push some of the stuff out.

Ah right, I'm guessing that they are more heavily protected on the larger CNC machines, plus there are numerous guards etc. From what I've been reading Ballscrews tend to be more accurate, certainly they are the only ones used in general CNC machines. Anyway, easy enough to create something that can protect the screw from debris yourself isn't it?

[skrubol]

Keeping debris out of the ways and screws is pretty important in any machine that needs to be accurate and/or last. Some sort of way cover is always used on 'real' machines, whether it's accordion/bellows style (usually only for Z axis on bigger machines, accordion folded covers laying flat accumulate chips,) or sliding (usually metal,) covers. X axis on typical milling machines don't need covers because they're basically upside-down, so the mill table acts as its own cover.

[Generic Default]

Yeah within the hobbyist DIY CNC builders, everyone uses cheap ballscrews and cheap linear rails. If I were building a cheap router to cut plywood, I would definitely go with these. They have almost no resistance to motion which means small steppers can push them at higher speeds. The huge downside is that all the plywood dust or tiny metal chips can get into them, and once the balls start rolling over that stuff there is no way to get it out.

All cutting in the SwissMak is done within only a few inches of travel, and cutting speeds during turning operations don't require much linear feedrate at all. I've actually tested resistance to contamination by cutting cast iron (notoriously bad for machines) and sprinkling the chips directly onto the leadscrews. This would instantly ruin a ballscrew, but the acetal leadnut makes a great wiper and simply pushes them out of the way. It's essentially like an O ring.

For machine covers on the SwissMak, I planned from the beginning on using a torsion-spring loaded roll of abrasion resistant fabric or sheet rubber. This acts like a very wide tape measure, always rolling itself up by retraction to keep the shortest length of sheet exposed and covering the ways and screws of the machine. Kind of like way covers or bellows except without any folding or sliding action. Ultimately this is unnecessary because the boxways and screws are immune to fluids and contamination.

[skrubol]

Ultimately this is unnecessary because the boxways and screws are immune to fluids and contamination.

The ways/screws themselves may be, but coolant will wash away your way oil.

[Krispee]

The Hurco's we use (like all CNC machines) have good covers, two or three together, each one folding under the other, and using a special seal that squeezes out the cooland and swarf. After a while though those covers get dented and warn and regardless of the the state of the seals let in coolant and swarf, and other than replacing them completely you can't do a lot about it. You should have seen the crap gathering around the main Y Ballscrew we found when we took them out one year, inches deep. For my own design I want to focus of the accuracy so am worried about using Leadscrews because they don't get ground, not from what I've seen anyway. I'll just have to design a wiper to clear them and make sure they are guarded.
I don't know much about kickstarter campaigns, hope that works out for you.

I have included a link to some wipers I just found that guard Ballscrews from contamination. Interesting design.
https://www.designworldonline.com/co...n-ball-screws/

[Generic Default]

skrubol, the boxways and leadscrew nuts on this machine don't need grease or oil at all. They run fine completely dry. Metal on metal contact needs lubrication but slippery plastics like acetal work fine dry, so long as the speed is low.


Krispee, I agree on the way covers. Many machines with boxways (even modern ones) are built without way covers because the boxways don't crunch chips, and the coolant gets pushed out by the lubrication system. Everyone else with sliding way covers just has to deal with long term contamination.

The screw wipers you linked to look pretty cool. As long as the sealing lip isn't abrasive on the screw surface they would probably help a lot in a dirty environment.

Ground ballscrews with double nuts, preloaded on both ends are the industry standard for the better machines. The cheaper machines are lacking in ballscrew size and often only use a single ballnut. Cheaper machines don't have heavy thrust bearing blocks either, take Haas and Fadal machines for example.

The acetal nuts on the SwissMak are 3 inches long; this effectively cancels any small variations in screw pitch. The only way to get machine accuracy over any large travel is to comp the ballscrews with a laser and store the data set in the machine control.

[CFLBob]

When you get your Kickstarter URL, would you post that?

[Generic Default]

Launched the Kickstarter at 12 noon USA West Coast time, April 20th 2018.

https://www.kickstarter.com/projects...ur-machine-sho

[ianagos]

Launched the Kickstarter at 12 noon USA West Coast time, April 20th 2018.

https://www.kickstarter.com/projects...ur-machine-sho

Hope it turns out well this is pretty neat.

[691175002]

Can you provide more information on your previous business selling 3d printer hotends? It is very hard to commit four figures to a first-time Kickstarter.

Also what is the overall machine size.

[TinyWork]

This project is very interesting. I like your use of materials to solve problems, and the modular aspect.

Would you be interested in building a custom CNC mill built for our specific needs (high rate of repeatability and accuracy) based on this one? It could give you a cash boost, if you're looking for one, and would give us a turnkey solution for our project.

PM me if you're interested and would like to know our parameters, specification and work envelope.

Thanks, Steve

[Krispee]

Launched the Kickstarter at 12 noon USA West Coast time, April 20th 2018.

https://www.kickstarter.com/projects...ur-machine-sho

All the best with this, hope you get your target.

[dharmic]

Would have been able to say #first if it hadn't been for an overzealous credit card fraud unit at my bank freaking out

For those of you thinking "Looks neat, but price too good to be true and who is this guy making them anyway", I've been watching him build it up over the last three years, watching him come up with good ideas to solve problems cheaply, try them, discard them and repeat until he found something that works.

He's not a Kickstarter fly-by-nighter, he's the real deal.

Sure these machines aren't going to replace a million dollar commercial mill turn. But for guys like me doing random prototyping and small production runs out of their shed or in a small toolshop environment? Hell yeah. Get on board, make it happen.

[The_chip_troll]

I fail to see the benefit of using acetal nuts and making the nuts 3" long, could you shed some light on that decision? What is the error rating on the leadscrew lead?

Agreed on your linear motion comp point, BUT even c7 screws with double nuts are pretty damn good.

[dharmic]

Advantages I can see:
1. Acetal nut is cheap, self lubricating, self wiping, debris tolerant to an extent.
2. 3" long means more strength but more importantly a chance to build a slight pre-load into the thread to really clamp down on backlash.
3. When the nut ****s itself eventually (as even a ball nut will if not kept clean and lubed) it's cheap and simple to replace.
4. The feed forces of the steppers will be the limit for torque transfer rather than the strength of the nut - ball nuts are probably overkill.

The OP has done plenty of testing with good results. Given the price of these machines I've been happy to back the kickstarter - if it ends up that the acetal nuts aren't enough for what I'm doing then changing later to ballnuts and screws doesn't look like it'd be a massive job.

[The_chip_troll]

3" does not mean more strength. The acetal nuts I've seen in the 1/2" range (I don't know what this machine uses) are typically rated for axial loads less than 150 lbs. That's not much when it comes to machining. If you're using a nut as a spring to "clamp down on backlash", then it's going to act like a spring when you push on it. Acetal nuts are great on 3d printers, but I'm skeptical about their use for this application.

We have yet to see "plenty of testing with good results".

I think this machine is super neat. The OP seems to know what he is doing and it's a beautiful product, looking forward to more info.

[CL_MotoTech]

Does CAM exist that is capable of running this machine that is in a feasible/comparable price range?

[skrubol]

Does CAM exist that is capable of running this machine that is in a feasible/comparable price range?

Fusion 360 is free for hobyists, startups and $1200/yr for version with 5-axis for companies making $100k+/yr.

[CL_MotoTech]

Let’s be real, Fusion can turn a fourth or fifth axis and then continue machining and it can do fourth axis wrapping. That doesn’t make it true five axis CAM, not in the ultimate sense of that operation anyway, and as far as I can tell you would want that for this machine.

[Generic Default]

The_chip_troll, the leadscrew on the production machine will be coated with either DLC or WS2 for a coefficient of friction below 0.1. The long nut increases load capacity, rigidity, and slows the wear rate. It's in a tight fitting pocket under each machine assembly so there is no way to move the nut relative to the metal slab. It's pulled against it with M6 screws. Even without being adjusted the backlash on the prototype machine was under 0.001". That's on par with rolled ballscrews.

CL_MotoTech, we aren't providing CAM software but pretty much any CAM software will output GCode that the SwissMak can run. If you were to program a 3 axis mill part, post the code, and run it on the SwissMak, it would mill your toolpath into the face of whatever you had clamped in the chuck or collet.

I know most hobbyists think machining is impossible without CAM software, but keep in mind all the demo parts in the Kickstarter video were made with GCode that I hand-wrote in a few minutes. Once we get to the point of having a conversational control with easy to use macros, people will be able to program difficult parts faster than it would be to set up the part in CAM. Still, we'll be doing Fusion 360 demos later on.