Atlas 10" to cnc guidance

aefriot

H-M Supporter - Gold Member
H-M Supporter Gold Member
Joined
Mar 30, 2019
Messages
10
I own the Atlas 10" lathe already. I plan to make small parts to prototype a gearbox for sub 10hp vehicles as proof of concept. I could use the lathe manually, but feel I am unable to get consistent results. As I understand, sub-13" swing machines should not be considered for heavier cutting. I do not require heavy or fast cutting as I only need accurate results no matter the time to make them.

Feeds are pretty slow on a lathe so I assume I will be able to use steppers (possibly closed loop) with a good certainty that it will be accurate. Ballscrews will be a must as accuracy is desired. One manufacturer stated rolled ball screw variance of 0.050/300mm (0.002/12"). Ground ball screw is 0.025/300mm (0.001/12"). I believe ground ball screw from a reputable manufacturer is good enough for prototyping as I only need to make sure the parts work, not last forever. Though, 0.050/300 (0.002/12") on each of two axes may bring my leniency on precision to bite me in the rear. Maybe my wallet is bare and I am justifying cheating out with rolled ball screws. The size of which I am not familiar with in the least. I see 1605 quite often with smaller lathes. Am I getting beyond that size and should look to 2005 for added strength? And will there be a necessity for different size ball screws on crossfeed and saddle?
As far as steppers, I was thinking something like 570oz-in dual shaft steppers to retain manual operation. I may even be able to find closed loop steppers in that size range.

Upon writing down just some of the components to make this cnc lathe, I have to wonder if I am making more work for myself than I need to. I could screw up many components several times over to get acceptable parts and still not spend more time nor money than if I built this cnc lathe. But you know what? I wouldn't have a cnc lathe when I was finished! Sometimes we do things because we want to and can...not that we should.
 
Just because you can doesn't mean you should.

I can think of many other machines I would start with for a project like this....

Have you made the parts in question manually with this machine? If so, what caused you to think that converting your current machine to CNC will yeild better results?

Honestly, if this is going to be a business your time might be better spent cultivating a relationship with a local machine shop that can run your parts on their equipment. I used to sell custom go-kart engine parts and had relationships with shops that could crank out the volume I needed in their off time for way less than I could ever accomplish.

I haven't done a CNC conversion, but I have built a CNC mill/drill from different pieces and I can assure you it's a great learing experience. However, what you may learn is that real CNC machines are an order of magnitude more powerful and rigid than you're ever going to see with anything home built.

Post up some pictures of what you have and list how much the components you're looking at will cost. You can also ask on the "can you make something for me" section of this forum if anyone who already has a CNC lathe is interested in your project.

Our very own @vtcnc has a project going

https://www.hobby-machinist.com/gallery/albums/cnc-lathe-build.735/

Maybe check with him or some of the other folks whe have gone down this road and documented it here.

Cheers,

John
 
I agree with John, The Atlas is not the best choice for a CNC conversion. I'd look at a Hardinge with a dovetail bed or similar machine as a donor. Or buy a used CNC lathe and upgrade the controls. There are a lot of these older CNC machines now on the market for relatively low cost.
If your main goal is to produce small parts, another approach would to get a turret lathe, again like a Hardinge, these machines were designed to crank-out production runs of parts, before CNC was a thing.
 
Just because you can doesn't mean you should.

I can think of many other machines I would start with for a project like this....

Our very own @vtcnc has a project going

https://www.hobby-machinist.com/gallery/albums/cnc-lathe-build.735/

Maybe check with him or some of the other folks whe have gone down this road and documented it here.

Cheers,

John
I have an Atlas 10". Two things come to mind when I think about converting it to CNC. 1) Terrible choice as the bed ways are not prismatic giving you good rigidity nor do you have good options for ballscrew mounting. It would really be a frankenstein project and 2) you will regret it as it is nice to have a manual lathe around for tooling projects or cranking out a simple part you will need. Put your time into caring for the Atlas and it will not let you down, but it will as a CNC.
I agree with John, The Atlas is not the best choice for a CNC conversion. I'd look at a Hardinge with a dovetail bed or similar machine as a donor. Or buy a used CNC lathe and upgrade the controls. There are a lot of these older CNC machines now on the market for relatively low cost.
If your main goal is to produce small parts, another approach would to get a turret lathe, again like a Hardinge, these machines were designed to crank-out production runs of parts, before CNC was a thing.
My CNC lathe build is exactly following the path as suggested by Eddyde. Hardinge chucker headstock and 7" dovetail bed for rigidity. OmniTurn XY stage for motion. DMM servos for motors. Centroid control. The project has taken a back seat to other priorities this year, but winter is here and I'm spending more time in there so this is going to get more love soon.
 
I hear you all. It does make sense once I hear why the lathe is not a good choice. I will look for a better choice as time goes on. I'll just clean up the Atlas for manual work. Then find a set of gears for it to make threads. should I choose.
 
Don’t get discouraged. Just keep your eyes open. Best bet will be finding an older CNC lathe with old controls that you can retrofit.

Especially if you plan on going into production. Commercial equipment is way more rigid and designed to run for years straight.

Check @JimDawson for inspiration and guidance on this.

John
 
Back
Top