Luke11CNC Build Log ,I'm so naïve

[Jonathan]

That then would just be a bodge to hide poor workman ship.

Or cheap ballscrews - hence why I think my measurement of the backlash might be misleading.

Plus it wouldn't preload the nut it would twist the nut.!

Maybe, maybe not. It depends what's twisting. If you assume the rails are rigid (not quite true for supported rails as they bend upon the support, but not much) then it must be the gantry in between twisting and any force not parallel to the X-axis is resisted by the X-axis linear bearings leaving only the axial force for the nut. On James' machine, clearly the weakest point on the gantry is the joint between the sides and 6" wide cross piece, so that and the cross piece will bend. The ballnut mount and gantry sides are solid so won't bend noticeably. .. etc it's obvious I don't need to tell you.

Not suggesting one should actually do this intentionally. The force ratings of the ballnuts is huge, so I think you'd get away with it. I'd be most worried about the linear bearings.

[JAZZCNC]

Maybe, maybe not. It depends what's twisting. If you assume the rails are rigid (not quite true for supported rails as they bend upon the support, but not much) then it must be the gantry in between twisting and any force not parallel to the X-axis is resisted by the X-axis linear bearings leaving only the axial force for the nut. On James' machine, clearly the weakest point on the gantry is the joint between the sides and 6" wide cross piece, so that and the cross piece will bend. The ballnut mount and gantry sides are solid so won't bend noticeably. .. etc it's obvious I don't need to tell you.

Yes the Gantry will more than likely twist at the point where the sides meet cross brace at one side but the Ballnut is attched to the side so will twist against the screw unless the screw roates the same amount as the gantry twists. Plus like you say the bearings will suffer the worst.

Either way it's wrong and not recommended.!!

[Jonathan]

Yes the Gantry will more than likely twist at the point where the sides meet cross brace at one side but the Ballnut is attched to the side so will twist against the screw unless the screw roates the same amount as the gantry twists.

No, as the gantry sides will not twist at the point where the ballnut is mounted. Looking at the machine facing the Y/Z plane from the top it's cross brace, gantry side, X bearings, X ballnut. The gantry sides will twist between cross brace and X-bearings, but not much at all if any between the X bearings and ballnut as the bearings fix it. The twist will be tiny as the sides are very thick/strong aluminium plate. The same is true for machines configured like ours - the gantry 'ends' have the ballnuts mounted roughly at the same level as the gantry, and the X bearings provide support between so there's nothing to twist the screw. If the X-axis linear rails were not there then yes the screw will definitely be twisted, but that's just silly anyway. Imagine resolving each of the forces, any force not parallel to X that could cause twisting must appear as a contact force on the X-bearings and thus prevent twisting as there should be no clearance in the bearings.
In either case if there is any twist on the screw it's going to be a tiny negligible amount and the extra force that would apply is surely small compared to say the force from screw whipping.

[JAZZCNC]

Don't matter how you try to flog it Jonathan it's simply wrong and not a good way to build a machine. . . Sooner or later the price of inaccurecy and miss alignment of precision components will have to be payed for.!! . . The degree of inaccurecy will be the determining factor to how long it takes . . . . BUT Like Death & Tax's there will be no escaping it.!

[Jonathan]

Don't matter how you try to flog it Jonathan it's simply wrong and not a good way to build a machine

I'm not trying to flog it at all. I was just pointing out then that it is wrong to say there is any significant twisting force on the screw, and therefore you're only increasing the axial force on the ball-screw which is the same as what would happen if you did it properly with double nuts and a spring. If it wasn't for the large force it places on the rails I see no reason not to do it. It is obvious that there are far better ways to attain low backlash. I mentioned it originally more to point out how it could lead to a false reading of backlash.

[JAZZCNC]

I'm not trying to flog it at all. I was just pointing out then that it is wrong to say there is any significant twisting force on the screw, and therefore you're only increasing the axial force on the ball-screw which is the same as what would happen if you did it properly with double nuts and a spring. If it wasn't for the large force it places on the rails I see no reason not to do it. It is obvious that there are far better ways to attain low backlash. I mentioned it originally more to point out how it could lead to a false reading of backlash.

Well It wouldn't be acceptable to me or a practice I'd employ on any machine with my name to it.!! . . . . So again we'll have to agree to disagree.

Oh you might want to checkout how proper double nuts are connected and preloaded.? Dont think you'll find meny springs around.?

[Jonathan]

Well It wouldn't be acceptable to me ... on any machine with my name to it.!!

Same

Oh you might want to checkout how proper double nuts are connected and preloaded.? Dont think you'll find meny springs around.?

I know, but proper double nuts aren't as good as two nuts and springs / belville washers are they.
When I referred to 'double nuts and a spring' earlier I was thinking of doing something like this as it's a simple way of implementing it:

http://www.mycncuk.com/forums/showth...ull=1#post8924

Especially with only C7 grade screws where the pitch error limits how far you can go with double nuts.