Ball screws & drive slots?

Advert

New Replies

Report Bug

Help/FAQs

Ball screws & drive slots?

This topic has 10 replies, 5 voices, and was last updated 26 July 2015 at 18:22 by Muzzer.

Viewing 11 posts - 1 through 11 (of 11 total)

Author

Posts
26 July 2015 at 12:43 #12514
IanT
Participant
@iant
Possible Atlas MF Mill backlash solutions
Advert

26 July 2015 at 12:43 #198010
IanT
Participant
@iant
Saw a Tom Senior mill that has had ball screws fitted yesterday, still a manual machine but now very smooth and no noticeable backlash.

My thoughts turned to my elderly Atlas MF Horizontal mill, that has lots of backlash in all axis. I'd been thinking about making new PB nuts, involving also making acme taps etc. From curiosity, I looked at the cost of ball screws just now. Some ball screw & nut combinations direct from China seem almost affordable (assuming they are not really junk) and might just do for a manual machine . I then thought on about what other problems could arise and immediately realised that the Atlas table has a power feed, which is enabled by a long slot milled in the acme feed screw. A fixed drive gear engages with this slot to provide power feed as the table moves up and down.

The slot clearly doesn't bother the existing PB acme feed nut but I'd have to mill a similar slot in any new ball screw to retain the exiting table feed. Would this cause damage to the ball nut or are the balls retained in such a way they would pass over the slot safely

Any views – or anyone done something similar.

Regards,

IanT
26 July 2015 at 13:25 #198013
Bazyle
Participant
@bazyle
Is the drive dog/gear on the shaft next to the feed nut? They often are. Perhaps the modification can mount the ball nut on the gear rather than be fixed.
26 July 2015 at 14:21 #198014
Muzzer
Participant
@muzzer
Hi IanT

I bought a set of 3 ball screws and nuts from a seller in China on AliXpress for my CNC conversion. They offer to machine the screws to your drawings for a few dollars extra which I took them up on. Having designed my own complete conversion in CAD, I was able to print out a set of PDF drawings which the seller used to make my set up from. This included my own keyways, threads etc and a variety of diameters and lengths. I was pleased to find that they were nicely machined and the dimensions were spot on. Definitely worth getting that done, rather than only getting half the job done and having to machine an unwieldy leadscrew yourself.

My ballscrew conversion ballscrews retain the original Bridgeport-style endings but are extended to provide somewhere to mount the toothed belt pulleys. On the face of it, I could even reuse the Align AL-500 power feed unit but in fact I will be fitting this to the knee screw. My Z-axis conversion acts on the quill.

The main fly in the ointment is that the ballnuts are somewhat larger than the original, especially if you opt for the double-length "backlash free" version.

Murray
26 July 2015 at 15:05 #198019
IanT
Participant
@iant
Just trying to visualise it Bazyle.

The keyed gear (which is on the left) sits in a slot that is part of the saddle & feed drive unit. A 'fork' moves another gear (from memory) in and out of engagement with it. But the keyed gear always revolves with the feed screw (because of the key) but can slide up and down it's length freely as the table moves. The feed nut is stationary, the body of the nut sitting in a hole in the centre of the saddle.

So the gear always revolves with the feed screw but the nut is stationary with the feed screw revolving within it. So I'm not sure how the two would be connected but I'm probably missing something blindingly obvious.

If a ball nut can work with a drive slot in the ball screw, then the ball nut just has to be held in/attached to something similar to the body of the existing PB nut. Apart from machining the ends of the ball screw (to fit the handles etc ) and that would be about it. The Y axis has no slot to worry about, so should be simpler.

IanT
26 July 2015 at 15:30 #198024
Les Jones 1
Participant
@lesjones1
Hi Ian,
If my understanding of Bazyle's suggestion is correct then the idea it to rotate the nut on the ballscrew and the screw itself would be stationary. One other point to remember is that the pitch of ball screws is normally coarser than a normal leadscrew so the resolution of the graduations on the hand wheels would be less. (Or you would have to make the graduations very close together.)

Les.
26 July 2015 at 16:08 #198027
Michael Gilligan
Participant
@michaelgilligan61133
Posted by IanT on 26/07/2015 12:43:17:

… I'd have to mill a similar slot in any new ball screw to retain the exiting table feed. Would this cause damage to the ball nut or are the balls retained in such a way they would pass over the slot safely

Any views –

.

Ian … In my very limited experience, the balls are not restrained 'radially' in the nut [i.e. they are located by the screw], and would most likely drop into the slot.

[ Happy to be corrected by those who know better ]

MichaelG.
26 July 2015 at 16:15 #198028
IanT
Participant
@iant
Ah, thanks Les – I thought I might be missing something.

Bazyle, I'm not sure there would be room to do this in the existing unit without major surgery but there may be another way to fit something similar. I'll have to strip it down to see what room is actually available & where..

I like the sound of getting the screws machined to size Murray. If the cost was reasonable, that would make things very much simpler. I assume that the keyways you mention are in the end of the feed screw (for securing handles etc) and not along the body (e.g. main length) of the ball screw?

Regards,

IanT
26 July 2015 at 16:48 #198033
Michael Gilligan
Participant
@michaelgilligan61133
Ian,

Reference my previous post … I believe this is typical.

MichaelG.
26 July 2015 at 18:22 #198043
IanT
Participant
@iant
OK thanks Michael – I thought that might be the problem.

That might rule this out as a (simple?) fix for the MF but I quite liked the idea of sending a CAD PDF design spec off to China for the custom ball screw set – and I have an X-Y co-ordinate table that might be suitable for some experimentation with steppers etc. Joins a long list though…

Thanks everyone.

IanT
26 July 2015 at 18:22 #198045
Muzzer
Participant
@muzzer
IanT – yes, just at the ends.This is from one of the drawings (for the X-axis). I think I paid $10 for the machining of the X and Y balllscrews. Can't recall if that was each or total but it was a bargain either way. You can't really machine keyways where the ball threads are.

I made a transfer arbor for my Z-axis ballscrew which I bought a few months before the X and Y so I could machine the ends myself. I also had to machine down the hardened flange on the Z-axis ballscrew. If you need to remove the ballnut, it could be tricky with some designs, so you should consider how to make and use one of these arbors at the design stage. I'm not planning to remove the X and Y ballnuts at this stage but that may change….

Unless you buy from the US, ballscrews tend to be metric and generally come in 12, 16, 20, 25 and 32mm diameters. The most common pitches are 4, 5 and 10mm, although there are larger ones if you look around. 5mm is the same as the original pitch on my metric Bridgeport clone and I could see no obvious reason to change it. I also downsized from 32mm acme thread to a 25mm ballscrew. If you look at the load specs, 25mm should be more than adequate.

BTW, if you are made of money, you could order from Misumi who make configurable ballscrews (and many other components). You can customise the design, download the resulting CAD file and get it made and delivered. No, I'm not made of money but it's an intriguing business.

Murray
Author

Posts