[Fastest1]

I have 2 different mills where the Z screw is visible. Both exhibit a little bit of whipping/movement no matter what the rpm of the stepper. It probably allows more sloppiness than if the screw had bearings on both ends. It would seem then the mount would have to be dead on and the screw would need to be parallel to the axis with no tolerance.
Also much better use of the indicator judging by the pic. Did you say .0005? Maybe you need to pull out a feeler gauge to see just how small that is. If I build something and its is anywhere near that close, it is going to fit. I just googled average human hair thickness. It averaged between .00254-.004. Looks like your talking 1/10th of that.

[Winnfield]

The wobble isn't just at the bottom, it is just largest at the bottom. A 1/4" wobble at the bottom of the screw means you have a 1/8" wobble in the middle.

[Winnfield]

If a 1/4" wobble at the end of a <2ft. long screw is due to something on the screw itself, you could easily identify that with your indicator.

[DRock]

Hmm that is true...I definitely wasn't measuring anywhere near 1/4" of variation at the end of the screw when I had it in v-blocks so I suspect something is still a little off. I am going to go mess around some more and see if I can figure it out.

Do people typically use a thrust with an AC bearing like this?

[Winnfield]

The problem is that there is no alignment that does not depend on the preload of the AC bearing, which changes at certain points in the screw rotation due to some other factor, hence the wobble. You need a bearing that will provide the alignment, no matter what is happening up above. A deep groove bearing will give you that alignment, and they can actually tolerate a considerable thrust load. I would try a deep groove bearing below to start out with, and be careful with the amount of preload you are applying. Do not simply use another AC bearing above the one you have - you can't use them without a spacer to regulate preload. Try this first, then if it doesn't straighten it out, you will need a new solution for the bearing above. I'd use another deep groove and put a gas cylinder on the head to lighten the load. Of course, this isn't the ideal solution, either.

[SCzEngrgGroup]

Do not simply use another AC bearing above the one you have - you can't use them without a spacer to regulate preload.

I don't know where these crazy ideas come from! That is absolutely, 100% incorrect! A/C bearings are MOST often used with NO spacers, particularly in machine spindles and supports for leadscrews and ballscrews. Probably the most common bearing configuration in machine spindles is a pair of back-to-back A/C bearings at the tool end, sometimes two pairs, and a simple radial ball bearing at the top end. This is also the exact configuration used almost universally for supporting leadscrews and ballscrews in ALL kinds of machine tools.

Precision A/C bearings are sold in pairs, with the races ground to provide optimal pre-load when they are installed back-to-back with NO spacer. Even the pedestrian A/C bearings sold by shops like VBX will generally have more than enough pre-load for machines like these when mounted this way. I've built several quill drives for Bridgeport mills using these bearings, and had essentially zero backlash in all of them.

You NEVER want to mount A/C bearings at opposite ends of a shaft or spindle, as thermal expansion will make the pre-load vary by enough to damage the bearings, if not actually break the supports. And using a single A/C bearing at one end makes no sense whatsoever, no matter what you do with the other end. A/C bearings should virtually always be used only in pairs.

A/C bearings are intended to provide BOTH radial and axial support, unlike regular ball bearings, which have minimal axial load capability. Deep groove bearings are OK, though not necessary, at the top end of a spindle, or the non-thrust end of a leadscrew, but they have no business being used at the bottom end of a spindle, or the thrust end of a leadscrew. Whichever end provides thrust support, the bearing at the opposite end MUST be ableto "float", or thermal expansion can, and will, bend the shaft and/or damage the bearings.

Regards,
Ray L.

[hoss2006]

You NEVER want to mount A/C bearings at opposite ends of a shaft or spindle, as thermal expansion will make the pre-load vary by enough to damage the bearings, if not actually break the supports.

Regards,
Ray L.

BS. This is the way the spindle bearings are installed on many mills here that have years worth of use with no failures what so ever.
These mills don't have the room to use double lowers like professional equipment.
2 AC's at opposite ends of the spindle are proven to work well and reliably on benchtop machines.
Hoss

[DRock]

So then would it be OK to just use another AC bearing instead of the thrust? I mean it seems like it would work because there would only be about .75" between the bearings but I could bore it out for less.

What would the ideal way be to mount the Z-Axis screw?

[hoss2006]

I use 2 AC bearings on my z axis and it's going great for over 2 years now.
Hoss

[Winnfield]

I don't know where these crazy ideas come from! That is absolutely, 100% incorrect! A/C bearings are MOST often used with NO spacers, particularly in machine spindles and supports for leadscrews and ballscrews. Probably the most common bearing configuration in machine spindles is a pair of back-to-back A/C bearings at the tool end, sometimes two pairs, and a simple radial ball bearing at the top end. This is also the exact configuration used almost universally for supporting leadscrews and ballscrews in ALL kinds of machine tools.

Precision A/C bearings are sold in pairs, with the races ground to provide optimal pre-load when they are installed back-to-back with NO spacer. Even the pedestrian A/C bearings sold by shops like VBX will generally have more than enough pre-load for machines like these when mounted this way. I've built several quill drives for Bridgeport mills using these bearings, and had essentially zero backlash in all of them.

You NEVER want to mount A/C bearings at opposite ends of a shaft or spindle, as thermal expansion will make the pre-load vary by enough to damage the bearings, if not actually break the supports. And using a single A/C bearing at one end makes no sense whatsoever, no matter what you do with the other end. A/C bearings should virtually always be used only in pairs.

A/C bearings are intended to provide BOTH radial and axial support, unlike regular ball bearings, which have minimal axial load capability. Deep groove bearings are OK, though not necessary, at the top end of a spindle, or the non-thrust end of a leadscrew, but they have no business being used at the bottom end of a spindle, or the thrust end of a leadscrew. Whichever end provides thrust support, the bearing at the opposite end MUST be ableto "float", or thermal expansion can, and will, bend the shaft and/or damage the bearings.

Regards,
Ray L.

AC bearings only provide radial and axial support when used in pairs against one another, or with both races supported, otherwise rigidity is compromised, and this is well known. The constant load placed on both bearings is impossible to regulate and is often too much for these bearings to tolerate, unless they are massive for the application, I suppose (which certainly isn't the case here). The reason I suggested using the deep groove bearing is because they WILL tolerate the thrust load here because the thrust load is basically equivalent to a light preload only anyway since the head weighs 60lbs. Unless you crash it, it is unlikely that you'll break the bearing used that way. Hardinge has used typical deep groove bearings like this in ballscrew assemblies for years.

You are obviously not reading what has been said before commenting. We are talking about using it in EXACTLY the configuration you are saying is inappropriate - with only the outer races supported, and no support for the inner races. I am aware that they are used without spacers when used in pairs, against one another. This is not what we are talking about, though.

Using them alone or apart, without inner race support (no spacer) presents a rigidity issue, as well as a constant load on bearings with a 15 degree contact angle in most instances. We are not talking about using the bearings up against one another - I am well aware how they work and that they are designed for this purpose. GMN says not to do it, SKF says not to do it, NSK says not to do it, so I don't do it.

[Winnfield]

BS. This is the way the spindle bearings are installed on many mills here that have years worth of use with no failures what so ever.
These mills don't have the room to use double lowers like professional equipment.
2 AC's at opposite ends of the spindle are proven to work well and reliably on benchtop machines.
Hoss

Are the inner races supported as well? If so, how do you regulate preload? I have heard tell of a spring being used, but haven't actually seen an assembly in person. I think Ray is saying they can't be used without the spacer on the inside races if the outside races are supported some other way, which is what I was trying to say - not that they CAN'T be used - just that they can't be used like tapered roller bearings.

[Winnfield]

So then would it be OK to just use another AC bearing instead of the thrust? I mean it seems like it would work because there would only be about .75" between the bearings but I could bore it out for less.

What would the ideal way be to mount the Z-Axis screw?

Are you a gambler? How do you know when it's tight enough? All it takes is exceeding the load rating during a change in direction, etc. and you risk the bearing failing and the head falling down. This is part of the reason that a predictable, controllable preload is important for AC bearings, not just for precision, but because they are simply designed to work in pairs. If you are going to use another AC bearing, make a spacer that is EXACTLY the same thickness as the difference between the outer races. You can trim it accordingly on a surface plate with sandpaper for the preload. Read up on it a bit - you really can't use them apart from each other, without a spacer.

[DRock]

Man I am getting confused...can someone point me to a website with pictures/diagrams that illustrate the proper way to use these things?

Thanks.

[SCzEngrgGroup]

Almost any bearing manufacturer will have a design guidelines document that will describe in gruesome detail how to properly use A/C bearings. But, what you really need to know is the most common, easiest and best way to mount them in this application is in pairs, back-to-back at one end of the shaft (ideally the motor end), as shown about half-way down this page:

Proper Installation of Angular Contact Bearings - April 2006 Newsletter

The outer races should be clamped tightly together by a bolted cover on the housing, and the two inner races should be clamped tightly together by having one of them up against a shoulder on the shaft, and the other clamped tightly to it by a lock nut or double nuts on the threaded shaft end.

The other end should be supported by a standard ball bearing, that is a sliding-fit on the shaft, so it can accommodate thermal expansion of the screw. It is providing radial support only.

[DRock]

Thank you! That makes much better sense now...however I think that is not my real problem.

The issue is that the ball nut mount that bolts to the Z-Axis carriage is not machined small enough to fit into the pocket on the back of the carriage. I just got everything bolted back together and I went to put the carriage on and I noticed it wouldn't sit flush against the ways and that it was being held up. I took it off and measured with my caliper and the ball nut mount is around .050" too big to fit in the pocket.

I believe that this is what initially caused my binding issues when it was close to the top of the travel and that it bent my screw when I was running it up and down. This is unfortunate because I might be out a screw and a lot of time if it is bent too badly on the motor side.

I am attaching a picture for reference but I am going to email Billy and see if he recommends machining the mount down a little bit and then trying it again to see if it works.

What do you guys think?

[Winnfield]

Well, it has to fit in there somehow. You might as well mill it down.

[DRock]

Yeah I am going to bring everything to work on Wednesday and machine it during lunch. I hope I can use the screw as is, the power drill I use to spin the screw wobbles a good deal so I hope the motor is going to stay coupled during some hard machining.

[usmarine21m]

Thank you! That makes much better sense now...however I think that is not my real problem.

The issue is that the ball nut mount that bolts to the Z-Axis carriage is not machined small enough to fit into the pocket on the back of the carriage. I just got everything bolted back together and I went to put the carriage on and I noticed it wouldn't sit flush against the ways and that it was being held up. I took it off and measured with my caliper and the ball nut mount is around .050" too big to fit in the pocket.

I believe that this is what initially caused my binding issues when it was close to the top of the travel and that it bent my screw when I was running it up and down. This is unfortunate because I might be out a screw and a lot of time if it is bent too badly on the motor side.

I am attaching a picture for reference but I am going to email Billy and see if he recommends machining the mount down a little bit and then trying it again to see if it works.

What do you guys think?

I had almost the same exact problem you did. The Z ball nut mount didn't fit in that recess..made it fit.. Put the two screws in and the ball screw was having binding issues. Then a little screw wobble, so I had to fix my bent screw with an arbor press. Realized my ball nut mount needs to be a little taller so it doesn't pull up on the ball screw when I tighten those two mounting screws.