[Stevie]

Hi Guys

Well i got a long ballscrew off eBay (53" 0.500" dia 0.500" lead 2 start) with a single nut (2 ball tracks)
However there is a fair bit of play in it; Ive not measured it but must be 3-4 thou
So as I got it for a good price i have 3 options (as I can't find who made it for a replacement nut; and i already re-loaded the nut with new balls)
option 1
Toss it
option 2
build it into a third lathe as the Z axis (I'll have to cut it down) and add some form of backlash compensator; this could be a weight on a thin wire rope hanging from a pulley
option 3
I cast a plastic nut with bronze dust mixed into the resin; I've already tried this and other than the shrinkage it will work (I had no dust in it for the test)
I can allow for the diameter shrinkage by wrapping the ball screw with teflon tape before the pour; but it's the longitudinal shrinkage that i can't dream up a way of allowing for
This nut will be following the ball nut with a spring to push the ball nut against 1 side of the threads; the ball nut will be positioned to thrust in the direction of cut; so the backlash is always compensated for (unless your backcutting)

Any ideas guys

[HuFlungDung]

You could try here for some oversize balls:

http://www.precisionballs.com/

Its difficult to estimate what size you'll need, but you might be able to come a lot closer than what you have. That's not saying that the internal ball contact angle will be correct if its a crude screw and nut, but, it may get you by.

[High Seas]

Option 4 - How about using a short piece, make it into a tap - and drill/tap deltin/acetal rather than make a cast plastic nut? I and many other here have taken this approach with very good results. :cheers: Jim

[Stevie]

Thanks guys

I was told oversize ball did not exist

[ger21]

Moglice. I saw a website before, where they bored out the nut, and injected moglice into the nut while on the screw, so you end up with a ballscrew without balls, and no backlash. But I can no longer find the site.

[Stevie]

you mean this

[ger21]

Yes. Give tham a call and ask about that application. I also heard on one of the Yahoo groups that they have packages a limited number of smaller quantity packages, as the price is a little high. I think it was around $50 for the smaller package, but don't quote me on that.

[HuFlungDung]

Yes. Give tham a call and ask about that application. I also heard on one of the Yahoo groups that they have packages a limited number of smaller quantity packages, as the price is a little high. I think it was around $50 for the smaller package, but don't quote me on that.

Oops, my bad (I quoted him )

[D Walck]

The website for Moglice is www.moglice.com. Hope this helps. d

[Swede]

If the ballscrew is rolled (it probably is) then the backlash that is present is normal. Rolled screws have natural lead variation, and if a ball-nut is loaded with oversized balls, as the nut travels in areas with tighter pitch, then the nut will begin to bind. If you look at Thomson-type rolled ballscrews, they eliminate backlash with spring-loaded double nuts back to back. If it was a simple matter to load oversized balls, then I suspect they'd do it.

Although thinking more while typing ( ) perhaps the hand labor associated with oversized balls makes it cheaper for them to do the double-nut trick, so perhaps oversized balls will do the job for you! It certainly won't be an expensive test, just labor-intensive.

[HuFlungDung]

Yes, the trick is to guess what size balls to get. 1/8" balls are about 20 cents each, but as Swede says, it would be a lot of trial to load the nut and try it the full length of the screw to find the size that won't bind up anywhere.

It is good to know that every other ball can be smaller diameter. This lets you get away with purchasing only half as many "working balls" as the nut really requires. The small balls act as spacers between the large ones, to help prevent skidding and jamming of the balls against one another (think of what happens when you let your front bicycle tire rub against your buddy's back bicycle tire )

[Stevie]

The screw looks ground
I'm still waiting to hear from the guys about oversized balls

[Swede]

Steve, I've seen a number of rolled ballscrews that you'd swear are ground... not all of them look like Thomsons with the oxide finish. If it is ground, the oversized ball trick should work nicely. If rolled, it should still work, although there may be spots along the screw where the required torque might go up a bit ("binding", but not in the sense of a dovetail way; it'll just require a bit more torque).

The reason I think it's rolled is the backlash. Ground screws with backlash defeats the purpose of grinding in the first place. It could be ground but just worn a bit.

Good luck, let us know how it goes. I think guys are interested in this particular fix.

[Stevie]

I have looked at the ball nut closely and I would say it shows wear
I still have no reply
Maybe they just ignore the small guys

[balsaman]

I would not bother with a delrin or moglice nut if you cant make the ball nut work. If you need to go to that measure, use the srew somewhere else and get an acme screw and a delrin nut.

Also, delrin makes a nice tight nut but the backlash is unmeasurable because the more force, the more it gives. No way you can compensate for it. At least with with the ball nut you can dial in .003 backlash and it will be .003 regardless of load.
Same thing for spring loaded antibacklash systems. Enough force and the spring will compress.

Eric

[Stevie]

Eric

It is possible to use the spring method
I use it on the first lathe; the spring is pushing away from the main cutting forces; it will be a factor if you try to backcut; but i don't do that
The nice thing is that when the lathe is told to move back in a rapid; i don't need to overshoot and return to the correct location; because the spring pushes the nut against the flank of the thread that will be taking the cutting force the slide is positioned perfect for the next cut
The down side; is the spring is very strong; it keeps the nuts in constant contact with the flanks; and this causes increased wear
The nuts on the Z slide are ready for replacement now (4 months of useage)
The cross slide nut has a teflon button to keep the screw in contact with the nut (pushes down)
This nut I changed tonight; but the old one was 7075-T651 (tough stuff)
Now it is bronze; we'll see how long it lasts

[balsaman]

This is true for a lathe I suppose. Most of my experience is with routers, where the forces tend to be equal in both directions.

Soon enough I will be converting my lathe, so I am enjoying this thread, and learning.

Eric

[HuFlungDung]

I don't really like the concept of the spring-loaded anti-backlash nut, just because of the rapid wear characteristics. It would be better to space the two halves (with shims) to find a reasonable backlash limit (say .001" should be possible), and then fasten the two nuts firmly together (face-shim-face) and run it that way. This is the concept of a double nut ballscrew nut and it works fine there. It should also be feasible with regular screw threads. In fact, I know it is, because my old Misal mill has such backlash provision built into its bronze nuts.

[Stevie]

I run with a 0.0002 backlash
I need it to be this small; and here's why i need almost zero backlash

Thats a 25mm AA gun barrel made on my lathe; those are 1/2mm on a steel rule behind it
The barrel is 1/350th scale and less than 4.1mm long the biggest diameter is 0.251mm; there are 19 0.08 deep grooves 0.08 apart just before the plastic starts; they are tough to picture
I need to make about 800 of these and the same in 1/200th; but the 1/200th I have to drill with a tapered drill to match the taper on the muzzle

[HuFlungDung]

Backlash is only a factor of concern when making direction reversals. Is there any real need to change directions in Z when cutting that part? Impressive for its smallness, I might add

I suppose your carriage might "float" a bit uncontrollably when taking rough cuts, but I think your tiny parts would forbid much in the way of abusive "roughing".