/diy/ - Do It Yourself

>>1537339
>▶
That is not a CNC lathe.

>>1537339
>how difficult/expensive would it be to throw on some stepper motors

It's easy, and < 100 dollars. A lathe is useless without CNC so go for it.

>>1537342
I was thinking of controlling the stepper motors with a raspberry pi and adding a coordinate display with some push buttons

>>1537339
If it's missing the power feed, how exactly do you plan on adding steppers?
Skip this lathe unless it's very very cheap and you have the room to have more than 1 lathe

>>1537344
>A lathe is useless without CNC
t. basedboi

>>1537438
Why must hiroshimoot swap S O Y for based

>>1537437
Adding an independent lead screw for the carriage would be simple enough.
Use one stepper for left and right, another one for depth of cut.
If cutting threads, use an optical encoder to track what point in the rotation the main drive/work piece is. Then have the computer calculate x&y to engage the work piece at the same point for every cutting pass. I'm thinking as long as I start the cut in the same spot and the stepper motors keep a consistent feed rate I should be able to cut threads

>>1537344
You're on the right track but you're not going to do AND cut threads for under $100 like >>1537344 suggests.
>I'm thinking as long as I start the cut in the same spot and the stepper motors keep a consistent feed rate I should be able to cut threads
If you miss a step or there's even a slight variation in turn speed from cutting, you're going to turn some pretty imprecise parts.

>>1537443
>>1537449
Replied to wrong post

>>1537437
He's asking $900 but I think I can haggle down

>>1537449
this guy did it
http://www.embeddedtronics.com/cnc_lathe.html

>>1537459
Cool, I say go for it

>>1537438
well they aren't useless, but for anything other than repairs they are completely fucking obsolete.

CNC is amazing and how you get all the funky shapes you want.

>>1537466
>well they aren't useless, but for anything other than repairs they are completely fucking obsolete.

Then why are all the lathes in the tool room manual (where I work), while the production machinery in the main area is mostly CNC?

>>1537472
probably because you're repairing stuff instead of doing production

>>1537438
>t. basedboi

t. non-CNC fucking plebian scum that doesn't know any better

>>1537459
How did the guy get it indexed? Far as I can tell he uses a hall sensor for speed sensing but doesn't go into detail. Maybe the hall sensor also indexes the spindle?

>>1537480

That lathe in the video looked pretty awesome, compared to my similar sized manual lathe.

>Maybe the hall sensor also indexes the spindle?

from a Mach3 forum discussion:

https://www.machsupport.com/forum/index.php?topic=5483.0

"Mach 3 is programmed to accept pulses into the computer to measure the spindle speed. There are two methods. One is a simple one pulse per revolution of the spindle. The other is several pulses per revolution (these can be specified). What you need to know, therefore, is the system your existing lathe/mill uses. This needs converting to a 5 volt signal, and then can be fed into your computer."

So it could use a single hall to monitor speed and track the index.

>>1537344
>and < 100 dollars
hahahahahahahahahaha. no

>>1537459
>this guy did it
now i'm wet

>>1537477
funny looking lathe boyo

>>1537480
hall effect rotary encoders exist but let me blow your mind, if he always knows what the speed is, he knows what the position is.
although admittedly integrating like this is fucking garbage but if the spindle speed is real steady its ok

>>1537502
>although admittedly integrating like this is fucking garbage

if you mean mathematical integration that's not how it's done. in this case with a single hall you get one pulse per rev, which is your index point and also allows you to calculate speed.

and if your spindle speed varies much during one rev you have all sorts of issues anyway.

>>1537502
https://www.machsupport.com/wp-content/uploads/2013/02/Mach3_Threading.pdf
mach3 uses a timing pulse to correctly synchronize when to move the Z axis.
So the hall sensor also indexes in this fellow's case.

>>1537459
>He's asking $900 but I think I can haggle down

Run away, thats not cheap. Thats the going price for a larger atlas lathe that isnt missing critical components.
Be patient, dont buy a fucking money pit.

>>1537443
>Adding an independent lead screw for the carriage would be simple enough.

The crossfeed mechanism on these is pretty shitty, also the split nuts are shitty and break a lot too. Trying to retrofit that stuff is not at all worth the headache, especially at nearly $900

You know the guy pulled the gearbox, lead screw, and split nut and sold them as they are the only thing worth money on the machine, now is just trying to sell the carcass to some retard who doesnt know better for market price.

>>1537546
who the hell can't make a damn split nut though?

>>1537561
Depends on how hacky and shitty OP want to do it.
Hacking garbage together and properly retrofitting a precision tool are two very different things.

In fact, turning and milling out a proper atlas split nut out of brass has been on my to do list. I know ill be able to knock it out easily. Someone like OP who doesnt have the tools or clearly the knowledge isnt going to pull it off.

>>1537568
hory shiet

this file has tons of PDFs on atlas lathe stuff too

https://grabcad.com/library/atlas-10-lathe-1

>>1537568
use bearing bronze

>>1537466
Yep. And as you know, a home gamer like OP will be doing mass production primarily of course, and never do repair work or one off simple parts.

>>1537608
" " home gamer " " is code for rich person that buys a lathe for fun, everyone else will want to turn some profits

>>1537624
What do you consider rich? I bought my lathe when I made $20 an hour. I just bought it to support my own hobbies and have fun. I certainly do not turn a profit from it. A lathe is not super expensive. They are usually less than a car. People even buy those for fun.

>>1537627

>>1537459
thats a ripoff

>>1537627
yeah okay fine, the point still stands though, CNC is amazing, it turns complexity into something much, much more manageable.

https://www.youtube.com/watch?v=FmmNRaKpBTI

>>1537624
>can buy a lathe for the cost of a video game console or a laptop
>"rich person"

Wow

>>1537659
>CNC is amazing, it turns complexity into something much, much more manageable.

And it turns a regular old hobby into something exponentially more expensive.

>>1537339
for a lathe with no gearbox or powerfeed, id pay whatever the scrap iron price was for it

>>1537662
>And it turns a regular old hobby into something exponentially more expensive.

It doesn't, unless you're unwilling to learns the ins-and-outs of it, which itself seems kind of odd for the types of people who would be interested in this kind of thing for fun in the first place.

Any manual machine can be converted for a few hundred. Yeah, if you're going for a ready-made, turnkey kit, those are going to be way more expensive, but that's only because they're marketed to those who just want to use it to make money, and don't care much about the process beyond what helps them make that money.

A power supply, steppers, drivers, a few extra bits and bobs, and a PC running LinuxCNC (plus some kind of spindle encoder for a lathe) will get you a perfectly usable machine for hardly any cost. The only caveat is that you have to be willing to do the work to get it all put together.

>>1537659
>>1537671
>>1537344
Stop CNC proselytizing you fucking nerds. You aren't happy if someone else isn't into your precious fag hobby as much as you are.

>>1537674
>ree! no one wants your newfangled complexity!

>>1537671
I have some experience in CNC, home and professional.
It seems to me like you havent actually looked into the pricing or true requirements needed to do the conversion. OP posted a sizeable 12" swing lathe. He is not going to get a single quality stepper driver for less than $100 that would actually be suitable for that lathe.
Its like you are looking at 3d printer tier steppers and drivers sets and think they are going to be acceptable.

Im actually curious now, go on and post some components that you think will get OP in the ballpark.

>>1537682
OP here
I was thinking of using this system because I want to use it in another project
https://odriverobotics.com

>>1537682

I've built a total of 3 CNC routers so far, half-assedly working through a mini-lathe conversion. Admittedly, I have a big advantage in that I do electronics for a hobby, and can build the drivers myself. But

Zero effort answer, literally one of the first hits on eBay:

https://www.ebay.com/itm/US-Free-Wantai-3Axis-Nema34-Stepper-Motor-Dual-Shaft-1600oz-in-Driver-7-8A-CNC/181566990556?hash=item2a463c68dc:g:jpcAAOSwzJ5Xe2MN:rk:6:pf:0

I'll admit, the Chinese drivers are VERY hit-or-miss (the first one I bought worked fine up until I blew it up, the second didn't work at all without modification, the third worked, but shittily), but that's one axis too many worth of kit and it's still under $400. You could easily do better by shopping around/buying direct from China. Throw in an old PC with a parallel port for $50 and a thrift-store monitor for $20, if you don't already have those laying around already. It's not necessarily impulse buy levels of cheap, but it's nowhere near the "exponentially more expensive" pricing previously claimed, and certainly doable if you can afford the machine itself in the first place.

A lathe will also require an encoder for threading, but, again, how much that will cost you varies wildly depending on how easy a solution you want. Could pay $1000 for a commercial, bolt-on unit. Or, in my case, pay just cents for some opto-interruptors and a transparency sheet with some lines printed on it.

>>1537697
>doesnt mention the new spindle needed
>doesnt mention the ball screws
>doesnt mention the custom mounts

But oh yes, after buying all that nice expensive stuff lets throw the shittiest cheapest power train we can on it and then be surprised when you crash the machine hard and break something.
That doesnt even touch how much tooling itself will cost.

Its an expensive hobby, CNC makes it exponentially more expensive. (that is, if you want a machine worth a fuck, not some toy tiered cnc routers or a mini lathe that cant even turn steel)

He would easily spend more on the CNC conversion and tooling than the actual machine, this is a reality for most people. Especially with him looking at fucking brushless stuff.

>>1537739

You don't need a new spindle, what even are you on about?

You don't need ballscrews, though they're nice to have. The regular leadscrews will work fine, as long as you remember that they have significant backlash and program accordingly. Mach3 and LinuxCNC have functionality explicitly to help compensate for this, even.

You're really gonna pay for custom mounts when the _entire reason you own these tools_ is ostensibly for that exact kind of thing? Really, brother?

Again, I'm not arguing that quality equipment is expensive, but good manual machines aren't exactly cheap to begin with. We're working on a budget here, not jerking ourselves off over sub-micron surface finishes. If it works reliably, what else do you really need for hobby work?

You sound like one of those guys who's been spoiled by the tools his workplace owns and is completely out of touch with what someone realistically expects when they AREN'T buying a machine to make them $10,000/month.

>>1537741
>You sound like one of those guys who's been spoiled by the tools his workplace owns

You could call it "spoiled", or on the other hand you can call it "experience". You know, the knowledge of understanding why you would want to control your spindle speed or have zero backlash for your various toolpaths.

> Mach3 and LinuxCNC have functionality explicitly to help compensate for this

Mach 3 is too expensive for you, also its to compensate for your very very minimal ballscrew backlash, not to repeatedly smash your brass nut back and forth into the leadscrew to compensate for .010-.020 backlash.

>We're working on a budget here, not jerking ourselves off over sub-micron surface finishes. If it works reliably, what else do you really need for hobby work?
> what someone realistically expects when they AREN'T buying a machine to make them $10,000/month.

What exactly do YOU expect?
If you are on a budget, and you are only doing hobby work, and you arent looking for production work, and you dont want high quality finishes why the fuck are you trying to CNC anything in the first place?
Its far cheaper, faster, easier, and more reliable to do things on a manual machine.

You are half assing the job, and make no mistake, you are going to get a very half assed machine.
Just wait and see how impossible its going to be for your lathe to turn a simple ball without ball screws and spindle control.
Your axis movements will not be able to control the SFM and your holding torque will not be strong enough alleviate the chatter. You are going to try and try to compensate in software until your cycle times are so fucking long and your finish looks like complete dog shit and it still wont be good.

You may think you can cut every single important corner, youll eventually learn how your time and money is being wasted.

>>1537764
Thousands of people started their business with half arsed cnc machines. You go cnc to free up time to waste it again with bad cycle times so you can get away with cheap machines. And it is cheap. Two jmc ac servos, psu and old pc may run you 250$. Ballscrews and mounting maybe another 150-250$. You could build the spindle encoder for zero $.

A lathe (for most parts) obviously doesn't even care for backlash. Just ignore it. Controlling sfm is just for not burning up your expensive tools when talking about hobby stuff.

I'd do it OP but only, as someone already said, for scrap price. 900$ is insane. Double that and you could probably already get a working industrial cnc lathe in similar condition.

>>1537783
that's true, blue hands guy is one of them.

https://youtu.be/GiJns7R68iY?t=119

https://youtu.be/tOKfFwla8dA?t=438

here he goes over the janky cnc converted mini lathe, lots of vids of the journey. He makes some spare cash making car parts

https://www.youtube.com/watch?v=TNYUairOrrk

>>1537764
>You know, the knowledge of understanding why you would want to control your spindle speed or have zero backlash for your various toolpaths.

Then you should also have the understanding of why it's not NECESSARY in a hobby setting.

>its to compensate for your very very minimal ballscrew backlash

But it can be used, to limited effect, to partially deal with the larger backlash of a traditional leadscrew. And, again, I'm not necessarily recommending this, because the right way to handle it is to program with it in mind. It's just another option, if you feel you can make use of it.

>What exactly do YOU expect?

I EXPECT the thing to make serviceable parts for HOBBY use. How many times does this have to be said to get it through your head? Absolutely NONE of what you're listing is a requirement to make usable parts, they just make better ones. Yet again, however, for HOBBY use, this extra precision and speed is completely wasted.

>Its far cheaper, faster, easier, and more reliable to do things on a manual machine.
...right up until you either need to make two of them, or the part has any geometry more complex than a face parallel or perpendicular to the spindle axis. I mean, for fuck's sake, a taper attachment alone would run you half the cost of a cheap conversion, and has nowhere near the functionality.

You are making the assumption that entirely modern methods and equipment are somehow required for this to be useful. If, at the minimum, it could only do what a human operating the dials could, it would STILL at least be good for small runs of simple parts. But the reality is that even a cheap setup on an unmodified or lightly modified machine can do far better than that. There's any amount of evidence for this all over YouTube and the likes of CNC Zone, none of which I'm assuming you bothered to look for because you wouldn't be here otherwise.

if you're dead set on doing this, look for a lathe with shot gearbox but with lead screw
or even better, a Shot CNC where you scrap the 30 year old controller.
That approach seems a lot better than hacking a striped down lathe and you save the parts and headache

checking my Used-market, i could get the first option for 500$ and the second one for less than 800$ so look for a better deal, might aswell increase you search radius

instead of the O-drive, i've seen JMC Servos look promising and are on the cheap end

>>1537339
look for the cnc manual lathe conversion kits. We did that at a machine shop I worked for and they work pretty well.

>>1537796
>Then you should also have the understanding of why it's not NECESSARY in a hobby setting.

Heres the thing, you have not made a single good argument as to why a half assed cnc is NECESSARY, or even a good thing for "hobby use" to made "usable parts"
Youve literally boiled it down to, it helps you batch make parts. How often are you batch making parts for a project?
Yeah, almost never. And on the rare occasions you have to make more than a few things, was it truly a net gain having spent time money and effort on it?
You could be honing your manual machining skills instead of learning CAM and CAD, or buying nicer more versatile tooling instead of buying cheap steppers to automate only the most basic or turning functionality.

I think you need to define "hobby use", because it sounds like to me you want to produce a bunch of parts and sell them, but are arguing under the guise of just getting "usable parts". And it sounds to me like "usable parts" to you are very basic parts that have terrible surface finish. Its like you are trying to convince yourself your half assed attempt to CNC is going to make you a ton of cash on the side.

I am a hobbyist, and for me the whole reason im doing CNC conversions is to get more functionality out of my machines. Not so I can try to automate very basic movements to try and make a bunch of parts nobody wants.

>>1538019
>Heres the thing, you have not made a single good argument as to why a half assed cnc is NECESSARY,

Probably because not once did I ever say it was. My initial statement was a contest to the statement that it was "exponentially more expensive" than manual machining, particularly within a hobby setting. My only claim was and still is that it's not impossible or even very difficult to get a useful CNC machine, relatively inexpensively, if you're willing to do some leg work.

Never did I claim it would be comparable to a 5/6/7-figure commercial machine, in terms of speed or part quality. Yet, your entire point is just an elaborate non-sequitur about how it won't be, when that was not claimed at any point. The only statement you ever made that actually required addressing would be whether or not such a machine could produce even a bare minimum of part quality, which is completely refuted by the fact that such converted machines exist and do exactly that.

The only thing I was trying to say before you sperged out was that it was that it wasn't an option to be overlooked, because the cost wasn't all that significant. In my opinion, the perks of easily forming tapers and not having to manually change gears (obviously not a concern if you have a quick-change gearbox, which OP doesn't), just by themselves, is worth it, never even mind the myriad of other reasons you'd want a machine with CNC capability.