Getting (ball)Screwed

D

DMS

Active User
Registered
I am in the process (long process) of converting my milling maching to CNC. All the electronics are sorted, but I am still using the stock acme screws. I ordered a short length of 1" ballscrew stock (enough for the Y axis) and 2 nuts from McMaster a couple days ago; they should be here to day, which means I have to figure out what the heck I'm going to do with them. They are bare stock, so I need to machine the ends for the bearing blocks and pulleys. Thinking about how to do that, I came up with the idea of making some soft jaws for my 3 jaw, and then threading them to 4tpi so that they bite on the ball tracks, rather than the outer surface of the screw. Anybody have any opinion on this? Also, for those who have made lathe soft jaws before, do they repeat well enough to re-position the screw stock if needed?

I would try the 2J collets that came with my lathe, but I think when the previous owner put things back together it didn't quite right; I haven't torn it apart to see what is wrong, but the runout is about 30 thousandths, so that's a non starter.
 
Boring the soft jaws should be fine for the job, repeatability will depend on the condition of your chuck. It is the same deal when replacing a standard ACME screw, the ends must be finish machined by the end user. I have fit a few ACME screws over the yrs, and I used aluminum soft jaws to grip the screw. If you turn the work around, best to check to be certain the runout hasn't appeared. If need be, true the jaws again. If the screw is long enough, best to use a spider on the LH end too.
 
This is the short section (24inches) so I don't think I need a spider. If this works, and I move to the X axis (52 inches) I will definitely need the spider.

When you did the acme, did you grip on the outer surface, or on the walls of the thread? I ask because I have heard that the outer surface of ball screws is not concentric with the rest (at least, no guarantee). I guess I can give it a try and see what I get.
 
When I did the ACME screws, I held the screws on the outer surface. Gripping the walls would be tricky at best. I wonder if you wrapped the screw with a aluminum sleeve to protect the screw, and using a 4 jaw to dial it in would be another approach. If the walls are the most accurate place to take a reading, then I would imagine that you would need to drop a ball in the groove, and indicate off the top of that. That could be big time consuming, and could lead to errors depending on how you went about holding the location of the ball.

One each job I had done with ACME screws, I found that I couldn't do any better than 0.003 on the OD for runout. On each occasion I had to machine the ends, and wondered why the screw was sent without aleast a few inches that was turned down, or atleast one end centre drilled. It does make it tough to get the screw dialed in to machine the end features. From what I understand ACME is dirt cheap compared to ball screw shafting, and you may spend some time experimenting with different ways of dialing it in. Perhaps the longest set up before machining you have ever done.

I can see the need, and the want to get this right. Though the times I had to machine the ends of ACME screws, I allways came to a brick wall, and had to figure the screw wasn't that accurate, and Im gonna grow old trying to get the runout better. In the end, I had only the amount of shaft hanging out of the chuck jaws to machine the features I needed. I also machined a aluminum puck that I slid onto the screw, with a OD that slipped inside the spindle about half way.
I also used a spider at the far LH end for final support.

If you do find a way that works for the ball screw, by all means post your methods, as I for one would be interested in how you approach this problem.
 
Last edited:
Just a thought, how I'd try to do it,,,?

The ball track is (in theory) the critical surface that should be concentric with itself over the full length, could you "reverse" the measurement by wires process, lay short sections of wire (something soft, copper or aluminium) in the ball track and clamp on those? If the jaws clamping on 'em are concentric (eg soft jaws bored for the job) and the wire is constant diameter the ball track *has* to end up centred, I'd think?

Just my ha'pennorth,
Dave H. (the other one)
 
I have heard of people doing this and then using collets to hold the assembly (basically you wrap copper or aluminum wire in the ball groove then shove the whole thing in the collet). I haven't tried it myself, right now I am trying to figure out how I am going to eliminate backlash, oil the nuts, and build wipers.

I decided to try going the cheap route, and buy parts off the shelf; I'm starting to wonder if I got myself in over my head ;). In any case, it will be an adventure.
 
You must log in or register to reply here.
Back
Top