How to adjust or get rid of backlash caused by ball screw support bearing (fixed end)

[jevs]

I need to get the backlash reduced in my CNC machine. The backlash seems to be in the support bearings of the ball screws. I do not see any way to adjust this out?
Attached is a picture from the old manual and also some pictures of the real machine.

I am not sure if there is a way to adjust these or if you have to buy new bearings? Even though this machine is old, it has not had that much run time and it is in very good condition. It was never used in production and only for teaching programming at a college.

I have about .0030 to -.0034 backlash in the Y axis and about .0013 to .0014 in the X. I need to remeasure to get more accurate than that, but that is about what it is.

It has been awhile since I converted this to AC servos from the old Fanuc DC servos it used to have, but last time I was in there, I did not see any way to adjust it, but maybe their is, or maybe I need to shim it somehow? Or....maybe there is nothing you can do with these bearings except buy new ones?

This is a Pratt and Whitney Tapemate C, which is basically the first Fanuc Robodrill I guess.

Let me know if any more information is needed.

Similar Threads:

• Newbie- Can Pillow Block Bearing KFL000 be used as END Support for ball screw?
• New Machine Build- small diameter fixed end ball screw support bearing designs
• Ball Screw Bearing support
• For sale: 2 ball screws, 2 ball nuts 2 fixed end bearing blocks and bearinsgs.
• Ball screw fixed support with 0.46mm backlash?

[Jim Dawson]

After looking at the drawing, my best guess is the A290-6040-X108 shown on the drawing is an adjusting nut. Try loosening the cap screw, and rotating the part. I'm guessing it is threaded onto the ballscrew shaft. Looks like the cap screw is a thread lock.

[jevs]

This is what I am thinking also. I think it needs a spanner wrench or something. However, I could swear I tried this years ago and had no luck. Maybe there is no more adjustment left?? If not, then what can I do?
New bearings, or shim it somehow?
It seems like I can recall that maybe that collar was bottoming out, which means I would need a shim to compress the inner races together any more if there is even any space left....which likely means the bearings could be worn too much?

I can't wait to see how much bearings will be, but at this point the only hope of a cross reference will be to get the bearings out and look for a number and/or measure them.

I am guessing the only way to tell may be for me to just tear it all down (if what I think I remember is correct and the collar would not go in further). I will scan through the manuals again, but I seem to recall no mention of how to do any of this (and I have all the original manuals). I even called fanuc and they do not list these bearings anymore and the old manuals they have do not say anything about adjustment.

[Jim Dawson]

If the nut is bottomed out, then I think new bearings might be in order, but if the bearings seem smooth, then I would add a shim. I bought one of the shim kits from McMaster. Nice to have on the shelf for a lot of different uses.

https://www.mcmaster.com/standard-washer-shims

[spumco]

After looking at the drawing, my best guess is the A290-6040-X108 shown on the drawing is an adjusting nut. Try loosening the cap screw, and rotating the part. I'm guessing it is threaded onto the ballscrew shaft. Looks like the cap screw is a thread lock.

Yep. You can see the split in the nut where it clamps down and distorts the threads to lock it.

The split looks bottomed out in one of the photos. If the bearings are OK and the locking nut is old/damged/distorted, it may have rotated slightly and unloaded the AC bearings a hair.

I think the washer idea is the ticket. And McMaster also has pin spanner wrenches (or pin spanners, if you're pedantic) which might fit in there. You may be able to cut the shim washer and slip it on the screw shaft between the nut and AC bearing inner race without disassembling the entire mess.

Just make sure the washer OD isn't so big it interferes with the outer race or cage (if there is a cage or seal).

If there's no adjustment left (inner races are touching), then a temp solution would be to shim between the AC bearing outer races. If you're buying a pack of shim washers, might as well get another pack for shimming the OD. Same caution applies - shim touches the OD but doesn't interfere with the cage or inner race (or shaft).

I say 'temp' solution because if you're adding shims to the outer race, then the bearings are worn and it's time to call up VBX.

-Ralph

[jevs]

I am checking the Y axis today since it has the most backlash. It looks to me like the backlash is in the ball screw/nut.
I can put an indicator on the face of the split collar which seems to adjust the bearing play. I can pull back and forth on the table as hard as I can and there is zero movement. I would think bearing play would show up by doing this?
Now the other thing I did was put an indicator on the side of the table and pull back and forth. Again no movement.

What I can do is rotate the screw back and forth some without making the table move at all and seemingly without the ball screw moving axially at the bearing collar. This amount that I can do this also correlates somewhat close to how much the motor has to rotate to take up the backlash.

[Jim Dawson]

From what you say, it does seem that the problem is at the ball nut. But what I don't understand is why you got no table movement if there is some backlash in the ball nut/screw.

[jevs]

Well, apparently the reason I get no movement is because it takes a bar to even move the table. I took the ball screw out.
I could not move the table manually with both hands and feet. It is a big heavy table, but do you think the gibs are too tight? I could twist the screw by hand and move it, but I cannot budge it with my weight pulling on it. A long pry bar will move it. I wonder how easy it should move?

They had that split nut on there so tight I had to use the air hammer to get it loose.
The way it is made, the nut does take the backlash out. I found a reference to it in the manual, but basically just one sentence and no procedure. I would think it would just be snug though and not torqued like crazy. I had to use a puller to get the Oldam off. It has some sort of a tapered washer in there that the collar presses on to compress it to the shaft.

Anyway, I got the Y all out, but now I am not sure how to verify the backlash in the screw with it out. It all moves too easily.
I also wonder how they get the backlash out. Is the only way to put bigger balls in? This is an old machine. Is there a chance I could replace the ball screw assembly with something better and more modern as far as backlash goes?

The bearings and ballscrew are NSK.

The nut of the ballscrew assembly has this on it:
NSK
W2805-202-HS
7NX-067
I believe it is a 28mm ballscrew but it is cut for inch pitch.


The bearings are 7204A with some other markings on them.

Keep in mind this is the original 41 year old stuff However this machine was never in production. It was used for teaching NC programming at K-State until we got it from the state sale a long time ago. I have not done anything with it except make maybe 6 or 7 simple parts and moving things around trying to get Mach3 and then Mach4 all programmed to work properly.

Now I am down to fixing this backlash and it will be ready to make parts.

[Jim Dawson]

The table does seem a bit tight, is is getting oil?

Nothing wrong with NSK hardware, very high end stuff. With the single nut there is no way adjust it except to put in bigger balls. You can buy them in 0.0001 increments. I did have a vendor located, but they seem to be out of business now. Your local bearing house may be able to help you. There are also a number of ballscrew rebuilders around.

In those couplings there is normally a set of tapped holes next to the bolt holes, screw the bolts into the tapped holes and it should press the coupling off of the collar and free the assembly from the shaft.

The split nut should be set to preload the angular contact bearings just enough that there is some noticeable drag on the ballscrew.

Looks like a nice setup, you just need to get it working correctly again.

[jevs]

Yeah, the table is getting oil. The ways stay wet with oil and I have it programmed to run the oil every time the X or Y motor moves. I also put a tape line on the pump to make sure it goes down. I have not used it for machining anything though since I started putting Mach 4 on there, so it has not moved much. I have spent a couple months now just setting it up and learning to computer program in Lua to make it do what I want Long process.

I wish I had an idea of how much pressure it should take to move the table. Seems like the only way I might know is to loosen the gibs up and then see if I can move it by hand. Is it possible it should be hard for me to move by hand? It is a heavy bed mill table. The whole machine weighs about 4000 lbs or better. I can twist the screw by hand and move it seemingly easy enough, but I would think I could move it without the screw in there and pulling with both hands using my feet to push off also.

The only holes in the oldham coupler are the 4 the screws for the collar go into. It came off okay with a puller though.

Is there a way or a link to some info on how I can test the ballscrew backlash now they they are out? Seems like there will need to be some pressure on things to show it....like the table I couldn't move by hand that was giving resistance before I removed it.

Is it normal by todays standards to have a mill like this with a ball nut this style? I have no idea what the original backlash should have been? Can it ever get close to zero with this setup?

[Jim Dawson]

Well the table is heavy, but you should be able to move it, once you overcome the stiction. Sometimes they are a little hard to get started, but normally move freely once moving. Every one I have worked with I could move the table by hand, so maybe the gibs are a bit snug.

It is not uncommon to have a single nut, that's what my mill has. I am going to have to work on mine also, X seems OK after I shimmed the thrust bearings, but Y has about 0.004 backlash. Not sure it that is ballscrew or thrust bearings.

The only way I can think of to test the ballscrew backlash is to anchor the nut and apply pressure to the screw. It's going to require some fixturing to hold things in alignment to test. The mill table would make a good base for a fixture, flat and nice T slots to attach stuff. I would expect that you can achieve zero backlash with the proper preload on the balls. A tedious job, but worth it in the end.

[jevs]

Does it normally take a lot of pressure to see the movement?
I could mount the ballscrew in my lathe and then put an indicator against the nut and push back and forth on the nut. Is this good enough to show the movement, or does it need a lot of pressure?

If I do see backlash, is there any reason to send these out to be rebuilt, or just take them apart myself, measure the balls, and get bigger ones? The nut moves smooth and there no corrosion or anything on the screws. I can clean them and do a visual inspection though.

Any suggestion on replacing the bearings or how to check if they are still okay?

[Jim Dawson]

The pressure should be at least equivalent to the force needed to move the table. Maybe use the tool post apply pressure, that should help keep things in alignment.

It can be a DIY project, I've done a couple. One trick is to use different size balls, one small and one larger, alternated, this helps keep them from getting jammed up but still accomplishes the goal of tightening things up. About the only check of the balls is to mic them and visually inspect.

Or you could just send them out for rebuild.