[Fortune7]

I watched Keen's (Threadexpress) youtube videos on 'Tormach Z - quick adjustments) with great interest and afterwards decided to check my Z axis linear accuracy. Using a Heimer (and checking with another DTI) I have around 0.1mm (.004") 'lost motion', so decided to adjust.

After following Tormach's procedure for both the gib and ballscrew bearing pre-load, I can't improve below about 0.08mm, or 0.0031". The specification is 0.035mm (0.0014") and the test sheet for my mill (s/n 70971) shows 0.001mm or 0.0004". Keen managed after his adjustments to get 0.001mm so that's what I'm after, or thereabouts.

I've been over the procedure at least 3 times for both adjustments and can't improve. I removed the Z axis stepper and substituted a 'shaft and handle' to feel the ballscrew rotational friction as I don't want to go over-tightening the pre-load unnecessarily. The problem I guess is 'what feels right' in these circumstances. The gib adjustment is 'certain' as the lost motion can be seen to increase on the DTI when it's too tight, so the 'sweet-spot' can be determined. The ballscrew friction, however is down to 'what feels either free or tight', and for me too subjective. With the Z axis stepper (and hence brake) removed and the gib properly adjusted, should the mill's head drop (spinning the ballscrew) under it's own weight? Mine doesn't, so is the pre-load too tight??

EDIT - Update: I again removed the Z axis stepper, coupler and loosened the two adjusting nuts so the ballscrew bearing friction reduced to allow the head to drop under its own weight, then reassembled and applied preload. No change, still no better than 0.1mm! Tormach state in the manual that the ball-nuts are pre-tensioned for backlash (not user adjustable) so the lost motion HAS to be in the ballscrew angular contact bearings, right? Maybe the bearings are shot, but in less than a year's light use??

Any experienced help would be appreciated!

Cheers, Andy
Oxford, UK

[nitewatchman]

Check the ball nut mounting to the headstock.

[Fortune7]

Thanks for the feedback. It turns out my error was magnified x2 as at the time I was using the RapidTurn so my Pathpilot was in 'RapidTurn' mode - where the (mill) Z axis becomes X, and measured diameter, not radius.

Switching back to 'Mill' the best I can now get for Z backlash is 0.04mm (0.0016") which is much closer to specification but not the best achievable. For now, I'll leave it there and try again another time.

Cheers, Andy

[Fortune7]

With no small amount of embarrassment It looks like my machine is, after all, well within specification! My problem was two-fold:

1. I was measuring 'Z' axis lost-motion when in Pathpilot's Rapid-Turn mode, so the 'Z' becomes 'X' and the DRO readout is doubled (for lathe diameter, not radius). Thanks, Keen, for putting me straight on that one!

2. Not satisfied that (back in 'Mill' mode) the Z lost-motion was acceptable, today I checked again using a sensitive DTI and Tormach's published procedure. 0.01mm (about 4 tenths) is what I'm getting, so it's as-per the original inspection sheet, for all 3 axes.

So why when I did the original tests with my metric Haimer in the spindle were the figures worse? I can only assume the Haimer has its own backlash but how this can happen when the probe's under spring pressure beats me. Mind you, I've made such a balls-up of this whole episode I'll just put it down to my own dumbness........

Duhhhhh

Cheers, Andy

[IMT]

Andy, thanks for posting and allowing everyone else to benefit from your "experience"