View Full Version : Is this an anti-backlash nut?

05-06-2014, 01:50 AM

Not the the aluminum coupler but the the dark little nut to the left of it on the ballscrew shaft.
If so how tight is it supposed to be? It was loose and causing all sorts of issues I'll detail in a video tomorrow but I tightened it a little, as much as I could without the the right tool, and it really fixed the backlash issue I started having. I'm wondering if I should get back under there and crank it down, I just don't want to damage anything.

05-06-2014, 08:12 AM
What does that part integrate/interact with? Does it just spin on that shaft, or is it threaded in to the block, or ??? If it is just a locking collar, it would be a very poor anti-backlash "nut", but could limit a really bad case of backlash. An anti-backlash nut would normally be two pieces that are working "against one another" to reduce or eliminate the backlash.

bill south
05-06-2014, 08:13 AM
That is just a coupling collar!
If the screw was slipping in it, the backlash would be horrible.

05-06-2014, 09:35 AM
That is the preload collar for the ballscrew bearing.


05-06-2014, 11:10 AM
So how tight should it be? I basically just snugged it up since I couldn't put a whole lot of torque on it without the stepping the motor then I tightened the set screw to hold it. The nut is threaded onto the shaft and when unscrewed backs out to the aluminum coupler. I can't really see in there very well but when tightened down it seems to butt up against something inside the housing like a bearing I guess. I need to check the Y axis one too.

05-06-2014, 12:14 PM
Loosen the grubscrew in the nut/collar, and tighten firmly, but do not put lots of force on it. Gently does it, away from the coupler then tighten the grubscrew.
There will be two bearings in the housing and the collar puts preload on the bearing, axially restraining the shaft.
The amount of preload must be greater than the cutting force it reacts against.
Need pictures of the nut on the ballscre to if it is preload type. (two nuts on the ball screw coupled together.
Be aware that the coupler of that type, often break due to fatigue, although they still usually drive but become a cause of backlash.
A made helical couplers to replace two that broke. Used a tilted 4th axis and a slitting saw to make the couplers.

Partially visible in top photo is an anti-backlash ballscrew nut. It is two nuts linked together.

http://i399.photobucket.com/albums/pp79/neilw20_2008/SX3%20operations/BrokenCoupling.jpg (http://s399.photobucket.com/user/neilw20_2008/media/SX3%20operations/BrokenCoupling.jpg.html)

Un-finished helical coupling

off the shelf one...
Google Image Result for http://www.magnova.com/torque/accessories/Helical%2520Coupling.jpg (http://www.google.com.au/imgres?imgurl=http%3A%2F%2Fwww.magnova.com%2Ftorque%2Faccessories%2FHelical%252520Coupling.jpg&imgrefurl=http%3A%2F%2Fwww.magnova.com%2Ftorque%2Faccessories%2F&h=303&w=415&tbnid=ZUFSIHjnATdPsM%3A&zoom=1&docid=yu05en1osp89YM&ei=lAxpU8j-CMmIlAXZnoH4Ag&tbm=isch&client=firefox-a&ved=0CFUQMygCMAI&iact=rc&uact=3&dur=413&page=1&start=0&ndsp=17)

05-06-2014, 03:15 PM
That is the preload collar for the ballscrew bearing.


The manual for Tormach's 1100 mill has a section on adjusting the bearing preload and that might be of assistance to you. If the bearing nut is like Tormach's, you'll need a pin spanner and those are available pretty cheaply or you make your own in a pinch.


05-06-2014, 08:13 PM
Thanks for the info everybody! The Tormach manual is great! It appears to have the same collar as mine so maybe this week I'll get out there and check all the axis collars to make sure they are still tight. I've only noticed the backlash on the X-axis and only recently but now I'm paranoid. Bad things can happen when that collar gets loose! You'll see in my next video later tonight or tomorrow morning.

05-06-2014, 10:23 PM
Nice find. If you can put up enough info for a DIY for the rest of us it would be greatly appreciated!!!!

05-07-2014, 04:22 PM

See post #15 for the video of what happens when this collar comes loose.

05-08-2014, 09:25 PM
I checked the Y axis collar and rechecked the X axis one as well. I didn't realize it but there are 2 set screws on the collar and I had only tightened one so the other one was loose. The Y axis also had one loose screw and one tight one. I loosened the set screws on them both and just made sure both collars were snug and re-tightened the set screws. I put an indicator on the table and measured the resulting backlash. Wow. The jog pendant's resolution when set to x1 jogs the axis 0.00005", in other words it takes 10 clicks to move the table half a thou. I could jog back and forth one click and pick it up on the indicator so I have 0 indicated backlash. It appears my backlash is now under half a tenth (0.00005" ). That is really damn good!

Before I did this and before the X axis collar came loose both the Y and the X had about 0.0004" of backlash. When the X axis collar came loose I was measuring about 0.0012" or so of backlash.

I didn't really even tighten the collars since I don't have a tool to do it, I simply just backed them off then re-tightened by putting a flat head screw driver in one of the grooves and snugging it down.

I'll show the DTI results in the next video.

05-08-2014, 11:31 PM
Sometimes, probably not in your case, a grub screw is against the shaft, and another on top of if locking it.
Always a good idea to remove screw, and you should see shaft, but sometimes another screw!