Ball lead screw backlash problems

I bought a previously retrofitted BP clone mill and I'm just getting it up an running with a new PC based control.

I'm seeing more backlash on the X and Y lead screws than there should be- .001" on the X and 0.004" on the Y.

The backlash is uniform on the full length of both screws, so it looks like the nuts are worn but not the screw themselves. So far I can't find any manufacturers names on the screws, and I can't see the nuts very well without taking the table off.

Is it possible to replace just the nut on a ball lead screw or will I have to replace the whole set to get rid of this backlash?

Thanks,

Paul T.

Put a dial indicator on the end of the ball screw, and Really verify it is not the thrust bearings. A machine repair tech told me that 90% of the time he is looking at backlash in a ball screw it is in the thrust bearings not the ball screw. Of course he works on Mazaks, etc. and they all have high quality ball screws in the first place. The ball screws in a BP retro fit are most likely HiWin.
You can also put the dial indicator on the table itself, and touch off the flank of the ball screw helix. So, as the table moves along, the ball of the indicator is riding on the ballscrew as it turns. This will tell you the lash in the ball screw itself, independant of the thrust bearings.

It is "possible" to add bigger balls, to tighten up a ball screw, but half the people here just lost their lunch from even reading that! It can be a super nightmare taking a ball nut apart,and getting it back together, and many will claim that it will never work AS good again. IOW a negative gain from the very attempt.

Another check is to see if the backlash is uniform at both ends and the middle of travel. Don't overlook the posibility of loosness, or flex in the Yoke.

One more word about thrust bearing, I have heard more than once that most thrust bearnigs are Killed by over tightening, on installation not from wear.

Hope this gives you some more things to think about. You need to know exactly where the problem is before you can correct it.

Pete

That depends on the ballscrew. If the pitch and center-of-ball diameter is the same, then one nut should be compatible with another...

However, it could just mean that it's not a pre-loaded nut, or that it's a double nut that needs adjusting.

See of you can get in there to get a look at it - or better yet, see if you can get a digital camera in there, and post some photos.

Oh, yeah - and I'll second everything that Pete just wrote, too...

I put fresh balls in the nuts on the X and Y axes of my Tree 325, and once I was finally clear on what I was trying to do (thanks to folks here at PM) it certainly wasn't rocket science or anything to be afraid of attempting.

The new balls did tighten up the Y axis noticeably, but there wasn't any improvement that I could detect on the X.

But then I'd suspect that many mills see a lot more travel on the X axis with Y being used to increment things, so the X screw and nut probably have more wear than the Y screw and nut.

cheers,
Michael

.001" is not a lot of backlash, you will have a hard time improving on that, Many, many new machines were shipped with .001" backlash. A little more or different lube might change that. Maybe you can improve on >004".

set a dial indicator on the table, and put the point on the end of the ballscrew ( cover removed)opposite of the drive motor. put grease in the centering hole and then a small steel ball will stick, so you have a good surface for the point
Move the table left and right.
Any reading is thrust bearing play.
Many CNC mills have a "backlash" setting that will allow you to overcome this error if you do not want to fix it.
Thrust bearings should produce no movemnet.
Ball screws on good mills have little (.0002) or no play when new due to preload.
Ball screws/nuts with more than .0009 play are considered "requiring rebuild" status for accurate work.
For home shoppers, look at the tolerance you want to be within, and also if you can dial out error with a dip switch setting...
Please note, as play increases like to .005, you will hear it..the machine will 'thumpt" when cutting circles or changing direction

Good Luck

Thanks for the suggestions fellas, it turns out that it is a problem with the thrust bearings.

The Y axis had the large 0.004" backlash. I discovered by tapping the end of the Y screw with a rubber mallet I could move the table through the backlash zone, so it looked like a thrust/preload problem on that screw.

After disassembling it, it became clear that the "semi-pro" that had installed the ballscrews had hacked things a little. The Y screw preload wasn't set correctly and the way the motor mount was designed you couldn't adjust it. I reworked it so I could set some preload using shims and got the 0.004" Y backlash down to 0.001", same as the X axis.

Increasing the preload any more didn't get rid of the last 0.001", but the machine has 6204 plain ball bearings in place for these thrust bearings, and when you put in ballscrews you're supposed to upgrade those bearings to 7204 angular contact type, which handle thrust loads more accurately, so the last 0.001" is probably due to these bearings. I doubt its wear in the screws as the machine hasn't been used that much and was only used to cut small thin pieces of aluminum.

The thing I was worried about is the "semi-pro" that did the retrofit didn't put any limit switches on the machine, so when I saw the large backlash I was afraid the machine had been run hard into the end of the screw, but it looks like that wasn't the case (needless to say I'm adding limit switches to the machine).

I'll leave it as is for now, 0.001" is good enough for the time being, and I'll upgrade it to 7204 bearings at some point down the road.

Thanks again for the help fellas, I feel a lot better knowing my Y ballscrew isn't shot.

Paul T.

If you ever do need to rebuild a ball screw, I think Rockford Ball Screw in Rockford Illinois rebuilds them.
Some may be successful at this, but my recollection was that the RBS folks were pretty reasonable. I had sent them about ten ballscrews - the kind with external ball returns - and they came back pretty good. This was maybe 10 years ago.

Ahh - good. Thrust bearings are easier to fix...