Seems very logical to me, Bob

… most of the caveats have been well-aired in recent discussions.

[but will almost inevitably be reprised here]

MichaelG.

NO. Not a good idea. The subject of ball screws not being a ‘magical’ improvement was covered in a recent thread wrt doing it to a round column mill.
Normally it is the nut more so than the screw. Is it ‘jumping’ yet, so worn that it slips? If not you have some way to go. If the idea of making a full new nut is daunting just make a short section two threads long. It doesn’t need to be accurate at all, just enough to thread on the screw. Add it to the current nut on the screw (off the lathe) until it acts as a lock nut. Do this at the unworn outer end of the screw. Release it by say 1/8 turn or enough to give 5 thou of play and join the nuts together with a bit of brass shim soldered across them.

Most work involves pushing the cross slide inwards so the main nut handles this. (except boring) The small nut is enough to pull it back and engages with less backlash.

I would be very wary of fitting a ball screw to a manual lathe as they are known to reverse under load if not held as it would be with a stepper.

Might be easiest to buy a length of Acme threaded rod and make a new screw.

Quietly amused to see my prediction coming-true

… most of the caveats have been well-aired in recent discussions.

[but will almost inevitably be reprised here]

Note … I still think Bob’s original idea has great merit.

MichaelG.

If you are going full CNC you can take a different approach, getting a cheaper? smaller (in order to fit) ballscrew with essential non backlash end bearings mounted off the back of the slide with only a (3D printed?) position indicator dial at the front. Note possible problem of ballscrew nut being too big to fit in current location.
The next decision would be how you CNC in the Z along bed direction – short range converting the topslide or heavyweight full saddle movement.

Too late for an edit, so here is the overview:

https://youtu.be/WyG4LOG8EfA?feature=shared

MichaelG.

.

On 19 July 2024 at 07:00 Speedy Builder5 Said:

[…] no comment about the cross slide moving due to pressure of the cut. […]

Although back-driving is a real risk in some situations, I think it is mostly hypothetical/trivial in cross-slide application with a pitch of about 5mm.

For a run through the required calculations, have a look here:

https://youtu.be/7w6lJm245w0?feature=shared

MichaelG.

.

I’ve CNC’d my S7 and IIRC I did find that with the 5mm x 12mm ballscrew fitted but no stepper it would back drive if I leaned hard on the slide.  Mine is a PXF S7 which has a larger tunnel for the feedscrew so it will take a 12mm one.  The nut had to be fitted “back to front” though as I couldn’t machine out a recess for it.  The Z feed is via the main leadscrew, too hard to fit a stepper on the topslide.  Anyway who needs a topslide when you have CNC?  Would be nice to have a ballscrew but I stuck with Mr. Myford’s standard screw – eliminating backlash isn’t so critical on Z.

Having “dual standard” electronic dials is a nice though but maybe not so easy.  Simple encoders just generate pulses – easy to arrange these to advance the slide a small metric amount every pulse, but need some software trickery to generate some subdivision of 2.54 pulses out per pulse in.  Of course all that is in effect done in a CNC controller where you could just switch to a different “imperial” profile.  Then you’d also get all the benefits of CNC like flexible screwcutting, wizards for things like tapers and ball turning etc!