Zero backlash?

I have just been reading another thread which mentions using ball screws(?) on lead screws to obtain zero backlash. Can anyone point me to an article which explains this? Seems to be a CNC derived technique, Googling doesn't seem to get the results I am after, probably because I am using the wrong terms.

Thanks,

Andrew.

Try " ball screw backlash" as your search criteria, should answer your question as to the methods and limitations.

Thanks KWIL, that search term worked very well indeed!

Regards,

Andrew.

Andrew, you will be after the double nut type if you want a preloaded zero backlash ball screw. These days the rolled thread ball screws from Germany are just as accurate as the ground ones. There is also varying degree of the amount of preload as well.

Neil

If you need adjustable backlash, as is likely in some applications, you can best do that by using two normal ball nuts, and you rotate them one against the other until you get the desired result. I have seen some people advocate shimming, but that is rather pointless as using the properties of a screw thread gives you infinite resolution of the adjustment. It is cheaper using two standard nuts than having one preloaded at the factory to your requirements.

In my case, way back in 1978, the nuts each had an external keway. We fitted one on each side of the moving item. with a simple flange bolted to each side, at least one flange having curved slots to allow angular adjustment. When rotated gently the nuts clamped themselves against the accurately machined and slightly counterbored (for centering) faces of the moving part, and it was easily possible to feel for the change between backlash and friction.

I got well under a tenth of a thou backlash and negligible friction on a 2mm pitch ground thread.

you will be after the double nut type if you want a preloaded zero backlash ball screw.

Not necessarily. Most ballscrew manufacturers can supply single nut, internally preloaded backlash free screw assemblies. These have two (or multiples of two) ball circuits ground with a slightl linear offset in the nut & the preload is obtained by ball sizing. Ballscrew manufacturers grade balls in micron increments & use these to get the required preload. We used these a lot in my last employment for CNC use, where they worked well and gave long, reliable service. If they lost preload due to extended use, a relatively inexpensive reball would get them going backlash free again. We initially used PGM screws, then Hiwin after PGM were taken over by THK, who dropped the internally preloaded PGM screw range.

We also used German "Star" brand backlash free rolled screws, which had adjustable backlash obtained by a varying the "crush" on the split ballnut housing. They worked reliably as well & were noticably cheaper than ground, preloaded types.

Bear in mind that backlash free, preloaded support bearings are needed with ballscrews to get a backlash free drive solution.

Nigel B

Hello Neil and Nigel,

I am not looking to install a zero backlash solution on any of my machine tools. I simply wanted to know how they worked.

I guess that even a "cheap" arrangement would cost half of what my machines cost in the first place and the other half would be taken up by the price of a DRO installation! I am quite happy working in the old fashioned way. In fact I am happier keeping old skills alive, than in jumping on the latest techniques. However it doesn't stop me from wanting to know how the new stuff works!

Thanks for your comments,

Andrew.

As with many engineering items, Chinse manufacturers can provide suprisingly cost effective solutions. I did a quick measure on the table screw of my FB2 clone - 20 diameter 3mm pitch trapezoidal screw, overall length around 700mm.

**LINK**

This would probably be a reasonable place to start and, at £38 delivered for a 750mm long screw & bearings, isn't too pricy ? This may not be totally backlash free (but it won't have a huge amount of backlash) & probably wouldn't be up to making sub-micron parts on a 24-7 basis, but I would be happy to give one a go at that price if I needed to replace a worn-out screw (or fancied doing a CNC conversion) - particularly as Pro Machine Tools list FB2 table screws at £156, with the nut another £156 !

Nigel B

Hm! Maybe not so expensive after all? I shall certainly consider such a solution when I need to replace worn nuts and lead screws.

I assume that wear will be very much slower than a conventional set up?

Thanks again,

Andrew.

I went bonkers about backlash. Eventually I sprung pairs of nuts together and held leadscrews in tension using Belleville washers. My larger mill is now zero backlash up to 1/4 ton while my little Roland desktop is sprung around 200 lbf .

Once you have achieved zero backlash there is no going back

I seem to remember that split lead screw nuts, for the purpose of reducing backlash, were roundly condemned by certain forum members. Perhaps someone could remind me why this is so?

Andrew.

Probably because the wear will be shared between the nut and the leadscrew, so the leadscrew will be unevenly worn, depending where it spent most of its time. Splitting the nut and eliminating the backlash will look OK until you move to another part of the thread where it will either be loose again or you will have forced your half nuts apart and stretched / bent the adjustment screw. You don't get anything like the amount of wear on a ballnut due to the way they work.

Murray

My ZX45 has the split nut and it is not really all that good. It can not be set to a low level of backlash like 0.05mm for example and hold that level for any length of time. Already after about 15 hrs of use, the backlash has extended to 0.2mm . I just can't be bothered to keep chasing it. With ball screws the real important part is keeping the seals in good order, so making sure that chips/swarf/dirt etc stay away. Making or coming up with a way to keep it dust free makes a huge difference to the life of the assembly, apart from the regular correct oil feed as well. Just need to find a way to do that on the ZX45 mill, especially on the X axis. There is also hybrid ballscrews that use precision ceramic balls, but they are way outside the scope of the home workshop. Preloading a pair of nuts with some form of spring system works well, and there are many options for this.

The problem split nuts a la Bridgeport etc. is that the clearance is closed up by distorting the nut. this leads to a poor fir and increased wear over time. Another problem is the use of bronze nuts. Bronze is a good bearing material when running against hardened steel with good lubrication at moderate to high speeds. It's crap when running at very low speeds with poor lubrication against unhardened steel. The end result of both these factors is growing backlash and worn feed screws.

A far better result for feed screws is obtained by using a cast iron nut with a fixed section and a separate adjustable section that is adjusted coaxially on a fine thread. A locking ring/nut is used to hold the adjustment. When oil is introduced to the space between the two sections by a one-shot lubrication system, it gets the oil to the point that it's needed.

The benefits of this setup are that:-

1. The thread flanks of nut and screw are always in even contact with each other, which increases the surface area in contact and minimises the wear.
2. The cast iron nut will wear in preference to the feed screw. Since the feed screw is expensive and the nuts are a simple machining job, this makes replacement cheaper.

Somewhere I thought I had a picture of the adjustable feed nuts on my Beaver Mill, but I can't find it at the moment .

I've often wondered if one could make a well-fitting nut and screw and pump air or hydraulic fluid into the thread so one has a kind of helical air hearing - a quick Google seach shows that such things exist.

**LINK**