Ballscrew Backlash

All info relating to the Denford Novamill CNC Milling machines

Moderators: Martin, Steve, Mr Magoo

Post Reply
JPac
Posts: 3
Joined: Tue 26 Jun , 2018 14:42 pm
Hardware/Software: Denford Novamill with LinuxCNC

Ballscrew Backlash

Post by JPac » Fri 16 Nov , 2018 17:08 pm

Hi all. I have measured backlash of 0.12mm on the X, all of this seems to be on the ballscrew/nut interface. This seems rather excessive..... Y and Z are around .05. My machine is from Jan 2000, and has a single ball nut on X.
Questions:
Am I right that there's no adjustment possible with the ball nut?
Is it possible to fit a new 1605 double ball nut without also replacing the screw?
Thanks

Madkite
Posts: 5
Joined: Mon 17 Aug , 2020 22:30 pm
Hardware/Software: Denford Novamill 3
Fusion 360 for cad/cam
Probably Mach3 but not built it yet.

Re: Ballscrew Backlash

Post by Madkite » Mon 04 Oct , 2021 19:28 pm

I to am interested with this.

I have tested the backlash on all 3 axis.
Z is about 0.025mm Y 0.03mm and X 0.05mm.
Having done extensive testing and fiddling with the X axis.

I can say it's not the preloaded thrust bearings. I tightend them and it mad 0 difference. I thought it might be magnetic backlash and a torque lag to the armature of the stepper.
But sticking a bit of long paper on the end of the screw you could easily see it's rotation respons to input. And it's rotation as the axis doesn't move.

So my last thing to look at is the ball screw. Either some kind of slack in the mounting, which I very much doubt.

Or it has backlash. Some ball screw nuts are adjustable. But I won't know till I get it out to look.

Has anyone had one of these apart to know?

Akorenika
Posts: 16
Joined: Mon 19 Feb , 2018 20:46 pm
Hardware/Software: Novamill VR milling version 2

Re: Ballscrew Backlash

Post by Akorenika » Wed 08 Dec , 2021 18:30 pm

I had backlash in X too, although a lot more than you at 0.4 mm. I discovered that the ball nut mounting was very poorly executed. They use a mounting plate that screws onto the ball nut. So far so good. When the plate is tight on the nut the ball recirculating tube has to point anywhere but down or the nut won't fit into its place. In my case , when tight, the nut was pointing in exactly the wrong direction. So the mounting plate wasn't screwed all the way on but left half a turn loose and just held with a small grub screw that licks the threads. Over time that all became wobbly. The fix was a shim between ball nut and nut mounting plate that put the nut orientation where it needed to be. The result was much improved although not perfect. I didn't see a way of adjusting the ball screw. Having seen how it all works I wouldn't be scared to source and modify a new ball screw. There isn't anything that can't be done on a manual lathe and the nova itself. Alternatively it might be possible to resurrect the old ball screw with fresh balls. I used to set up ball screws in Formula 1 which were used in inertial dampers. We were able to source ball bearings in 1 micron increments for that exact purpose. Fiddly but doable.

Post Reply