[N4NV]

I am looking at converting a watch lathe so I need really small ball screws. What is the smallest practical ballscrews (with ball nuts) available?

Thanks

Vince

[LeeWay]

I have seen some nice tiny ones on Ebay at times.
I bought some precision acme screws with antibacklash nuts that seem pretty nice. They are 3/8" diameter.
I haven't actually used them yet, but may on a small build eventually.

[Jason3]

Hi Vince,

Sounds like a great project! Will there be a build thread?

What size ball screw would fit? I know Steinmeyer were offering 3 mm dia. ballscrews a few years back. From what I've seen, 10 mm is the small end of commonly available screws, but 6 and 8 mm can be found too. I had a couple of 8 mm ones, they were certainly nice and compact, and didn't feel too fragile... I threw them in with a lot of surplus used ballscrews I sold, otherwise I'd have given them to you

Best regards,

Jason

[N4NV]

Hi Vince,

Sounds like a great project! Will there be a build thread?

What size ball screw would fit? I know Steinmeyer were offering 3 mm dia. ballscrews a few years back. From what I've seen, 10 mm is the small end of commonly available screws, but 6 and 8 mm can be found too. I had a couple of 8 mm ones, they were certainly nice and compact, and didn't feel too fragile... I threw them in with a lot of surplus used ballscrews I sold, otherwise I'd have given them to you

Best regards,

Jason

As always, there will be a build thread (I have 4 others). I am looking for something close to 5mm, 6mm might do. The total x travel would be about 1", y 2". Very little load only light cuts. My Levin spindle is accurate to .00005" runout. I have turned some watch parts by hand that were .12mm in diameter, so this thing would have to be accurate. Still need a source for ballscrews.

Vince

[BobWarfield]

Vince, I hear you on the accuracy requirement. If you can find proper ballscrews, more power to you. Don't have anything to add there.

But, if you can't, consider an alternative. The forces involved on a tiny lathe are pretty small. Delrin nuts on an ACME screw such as what Widgitmaster uses on his router might actually work pretty well for you. The Delrin springs back after the cut, which pretty well takes out the backlash. The material has a built in lubricant (Teflon), so friction is low. The disadvantage is that enough force will either deform it, or wear it out really fast, but you don't have that problem.

You might take a gander at how Widgit approaches making them, or even get in touch with him about the backlash. He could certainly measure what he gets from one. You could make a neat little screw and nut combo up really quickly to try it too.

You could also make a preloaded double nut system out of Delrin on Acme pretty easily that would further remove any residual backlash.

Cheers,

BW

[N4NV]

I am using a Delrin nut on my Taig lathe CNC conversion. I was hoping for ball screws because of the forces involved. The steppers for this project will have to be pretty small. There is quite a bit of friction in a Delrin nut if you have it tight enough to prevent backlash and at those forces, I would have to build it much heavier than I would like to prevent an twisting.

Vince

[N4NV]

Bob, after your suggestion I looked into acme screws a little more and found that Nook has one that will work for me. It is 1/4-20 thread with plastic nuts. With my 200 steps per inch steppers, even without microstepping I will have an accuracy of .00025". The screws has an accuracy of .0003" which is better than rolled ball screws. The plastic nuts have a backlash of .006" so I am going to cut them in half and put a spring in between the two halve which will give me zero backlash with the light loads I will be seeing.

I bought some very low profile linear rails from eBay that will work for the carriage and cross slide.

It will be an interesting project, I just hope I can work on it along with all the other projects I am playing with.

Vince

[rowbare]

If you are going the ACME route, have you checked out http://www.dumpstercnc.com/ for anti-backlash nuts? He doesn't have 1/4-20 but he does have 1/4-16. That would still give you great resolution.

[N4NV]

If you are going the ACME route, have you checked out http://www.dumpstercnc.com/ for anti-backlash nuts? He doesn't have 1/4-20 but he does have 1/4-16. That would still give you great resolution.

I already ordered the parts from Nook. I looked at DumsterCNC's website and what they are doing with their anti-backlash nuts is close to what I had in mind. The nuts from Nook were $28 each so they were a little more, but I will end up with a system that is closely matched.

Vince

[jdr]

Just read your postings.

You might want to try Starrett micrometer heads. Great for positioning small items. Very easy to mount.

Jack

[N4NV]

Just read your postings.

You might want to try Starrett micrometer heads. Great for positioning small items. Very easy to mount.

Jack

I have an XY positioning table that I purchased to convert that is set up that way. It just did not work out the way I wanted so I am going in a different direction.

Vince

[zephyr9900]

I bought some very low profile linear rails from eBay that will work for the carriage and cross slide.

Vince, I'll be interested to see the mechanism you design. I built an X-Z mechanism for my carriageless Feeler lathe from two THK KR20 actuators, with size 10 (9mm tall) Thomson profile rails with preloaded carriages (overall height 15mm) in parallel with the slides. But that is a 4.5" center height.

Levin, that's a fine lathe. 8mm? How low of profile are your rails/carriages?

Randy

[N4NV]

My Levin lathe is 8mm. I has a 2" spindle center height. I modified a Taig cross slide to fit on it by machining it thinner. I got to looking at it and decided I should be able to CNC it with some small steppers I picked up. The Taig cross slide I am using has too much back lash and too much friction for what I want. The rail/carriage height for the rails I found are .4" tall. The Acme screws I am getting are .25". If I hang the nut off the back of the carriage, and machine a little groove in the carriage with a ball end mill, I should be able to keep the entire assembly under .825" per axis.

Vince

[Riceburner98]

I see you're all set with the parts but for anyone else dead-set on micro ballscrews, I happened to be browsing McMaster for stuff and saw they have 6mm x 1mm ballscrews / nuts, accuracy 0.210 mm per 300 mm, and .05mm backlash. $173 for a 250mm set though! Yeeha.. And you'd have to buy 2 sets in order to get 2 nuts if you wanted to spring-preload. Yikes.. Also -> http://www.directindustry.com/prod/f...3-229349.html# ( http://www.faulhaber-group.com/servl...657&kdid=40929 ) with some 4.25mm screws, but no cost listed, so I can only imagine... 5uM per 2" accuracy, 0-5uM "axial play", zero backlash on request.. $$$$$$

[N4NV]

Thanks for the heads up, it's nice to know they are out there. My precision ACME screws arrived today. Accuracy is supposed to be .0001" per foot. Since my total travel will be about 1" of X and 2" of Z, they should be accurate enough. By cutting the nuts in half and installing a spring, I can get zero back lash. I ordered another ball slide today. By putting two trucks next to each other I can put the screw between them and keep the height down. The lower I can make it the better chance I have of fitting a decent tool post.

Vince

[Riceburner98]

I was wondering the other day whether it would be possible (or even likely) to have Acme screws more accurate than (at least rolled) ballscrews.. I figured the rolling process is really just mashing over metal to form the threads (well, in a controlled manner of course!) but the Acme threads I would assume are cut with a die. Good die = same threads until the die wears out.. Definitely be following your progress, those screws sound promising.

[zephyr9900]

The lower I can make it the better chance I have of fitting a decent tool post.

What kind of work will you be doing with this lathe, Vince? Faceplate/pot chuck work or collet work?

I ask because our projects actually seem pretty similar to me, other than the difference in scale. My Feeler (Hardinge FSM-59 clone) is a carriageless lathe, just a precision headstock and bed. It has a 4.5" center height. Your Levin has a 2" center height. A little over 2:1 scale factor.

I bought my lathe with the specific intention of doing collet work on it (small-scale model railroad wheels). The mechanism I built has travels of 1.63 x 3.60 inches, a little under 2 times your planned travels.

But my long travel is in X, because I arranged my mechanism as a gang tool plate. I placed the top of the X slide/toolplate as high as I could and still withdraw a collet (5C in my case) from the headstock, for maximum stiffness to take advantage of the spindle's accuracy.

I don't intend to hijack your thread, but here's one picture of my mechanism as food for thought (again, not knowing whether you plan mostly collet work or not...) I haven't bored and split the center toolpost, which will mount a drill chuck, yet. Obviously, the rear toolpost is the turning tool and the front toolpost dedicated to a parting tool. But all three toolposts are doweled to the toolplate and could be replaced by other customized posts.

Just food for thought...

Randy

[N4NV]

My setup will end up looking a lot like yours except for the gang tooling. Also, due to height restrictions, my Z axis motor will have to hang off then end of the lathe bed. I will be doing mostly collet work with 8mm collets. I would like to make watch balance staffs when I am done, which are about 1mm at their greatest diameter and .12mm at the pivots.

Vince

[N4NV]

I will post the build of this lathe in the proper place, I just wanted to show the conclusion of my search for a small ball screw and how I intended to use it.

The attached picture shows the basic layout of my Z axis. Besides the mess of my watch bench, in the picture you can see a Taig cross slide I made to fit the lathe. I had do some special machining to get it low enough to fit under the spindle. There is also I micrometer XY slide I put together. I have two ball slides that are going to be the Z axis. In the bottom of the picture you can see a spare lathe base that I will use for this conversion so I don't have to drill holes in my chromed Levin bed. I am pretty good at hand scraping now and will make sure it is better than .0001" flat.

In between the ball slides you can see the precision ACME screw I got from NOOK. It was cheaper than a ball screw but has better tolerance than most rolled screws. NOOK guarantees better than .0001" per foot. It is 1/4x20 thread. The little stepper at the end of the screw is 200 steps per revolution. This will give me an accuracy of .00025" per step. With micro stepping it will be ever better. I purchase 4 of these little stepper from eBay for $25 total with the encoders! The guy was selling the encoders and the stepper were something he threw in because he did not want to take the encoders off.

I also got some plastic ACME nuts from NOOK for the screw. I plan on cutting them in half and installing a spring between the two halves so I can get zero backlash. The loads on this lathe will be so light that it won't take much preload. The X cross slide will look similar. I have one more ball slide that I will cut in half to get me the X direction. The total X movement will be about 2", it looks like I will get almost 5" on the Z though 2" would be more than enough.

I hope to be able to make custom balance staffs when it is all completed. A balance stall might be 3mm long, 1mm at it widest, and have pivots that are .12mm.

Vince

[zephyr9900]

Sacre Vache! That's going to be one sturdy little mechanism! No wonder you needed the skinny leadscrew. Looks great so far!

Randy