[inthebay]

Hey Folks,

So, I managed to bugger up my X axis ball screw and am searching for a replacement. In the meantime, I've ordered some 10x30x14.3 double row angular contact bearings and some spherical roller bearings of the same OD/ID. Now, I'd like to redo my X and Y axis to eliminate the 3-4 thou of backlash I was seeing when the table was loaded up. So, that begs the question: What's the preferred stack of bearings for constraining my X axis? I've seen a lot of variations, and my old one used the stock bearings one on either side of the end plates. I'm thinking I put a thrust washer on the outsides of the plates, under the locknuts, and put the double row bearings on the inside as only axial support.

How have you done it? Have you had any problems? Is there a best practice? How do you deal with thermal expansion of the ballscrew?

[Zenji]

Sounds like you've misaligned ballnut with the ball screw, i'd relocate either the endplates or the ballnut.
A finger type dial indicator a test bar made from scrap should put you onto the cause.
By the way, i'm from australia and the term 'bugger' is used for cursing in relaxed conversation and not regarded as harsh, but in canada would the term not have a far harsher meaning, possibly homophobic one?

Cultural difference, dont shoot me please.

[inthebay]

Sounds like you've misaligned ballnut with the ball screw, i'd relocate either the endplates or the ballnut.
A finger type dial indicator a test bar made from scrap should put you onto the cause.
By the way, i'm from australia and the term 'bugger' is used for cursing in relaxed conversation and not regarded as harsh, but in canada would the term not have a far harsher meaning, possibly homophobic one?

Cultural difference, dont shoot me please.

Nope, I say buggered up as akin to screwed up. Actually, I was investigating my .0035" of backlash and when I slid the table off I lifted it a bit and bent the ballscrew. My fault entirely, but now that I dialed it I know its not the ballnut, but the bearings that were the problem.

I ran into this interesting graphic. Attachment 313216

[inthebay]

Bump. Can anyone show how they have their bearings arranged for the fixed and floating end?

[Upnorth4]

Interesting that this is a critical piece of the puzzle and there isn't an outpouring of support.
I bought a "kit"!from automation technologies.
The electronics are EXCELLENT, nice box, professional wiring and connectors, all connections made and dip switches set. I had the steppers moving the right way with Mach3 in literally under 30mins having never touched Mach3 before.....

BUT:
The mechanicals are CRAP. Sandblasted aluminum ( who does that?!?) they tried to cover BAD machining that looks like a beaver gnawed the parts out of scrap....
NO instructions at all
To the point there is no good way to set up the preload in the X and Y axis. Just a crappy nylock that is supposed to hold the preload.
After the first circle I tried to cut came out like an egg I stripped the mill down and found that the lock nut that preloads the y axis trust bearing bottomed out before loading the bearing and left me with a good .010" backlash.....
Given that this is where the force of the ballscrews is transferd to the dram you would think that the mounts would focus heavily on this area,
I have yet to see much info on this issue.....

[mactec54]

Nope, I say buggered up as akin to screwed up. Actually, I was investigating my .0035" of backlash and when I slid the table off I lifted it a bit and bent the ballscrew. My fault entirely, but now that I dialed it I know its not the ballnut, but the bearings that were the problem.

I ran into this interesting graphic. Attachment 313216

What you ran into would be not much good at all, if you have, the double row AC Bearing that is all you need, it has to be configured correctly, and you will be good to go

The thrust bearing in the drawing is a waste, the shoulder would go up to the Back of the Bearing, and the lock nut on the front locks everything together

[arizonavideo]

AC bearings have a direction.

In the drawing I see no reason for the thrust bearings. All that is needed is two AC bearings properly loaded.

A simple way is to use two washers. One for the inside race and one for the outside. Make one washer .015" thinner than the other and that will give you ABOUT the correct amount of preload.

If you look at the bearings inside race one side is thin and one side is thick. The standard way is to have the thin side face in and for the thinner shim to be on the inside race.

If you do it this way you bolt the bearings to the screw and then the OD of the bearing is clamped to the housing.

I do it different for my 704 kits but you need to cut a double sided pocket in each bearing mount. Both ways work.

[Upnorth4]

Dredging up an old post because I can’t find a good definitive solution to the ball screw bearing mounts that seem to be coming back for me.

I like Arizona Video’s kits, if I didn’t already have screws I would go with their kit for the hardware.

I am curious though, are angular contact bearings needed / are they any better than a thrust washer for this application.

It would seem that there is little to no radial load. The stock thrust bearing in a decent mount may be enough?

Of course if I were needing bearings I would purchase AC bearing for this but I already have thrust bearings and can’t see it being worth it to get an AC bearing for radial load that isn’t really there….. any small amount would be dealt with the stock thrust bearing and or a good fit on a proper mount…..

I 3d printed a Y axis mount and it seems to be working well, if anyone is interested I can email the .stl files. In the long run I’ll machine the mount out of 6061, but for now it looks okay printed.

AC bearings have a direction.

In the drawing I see no reason for the thrust bearings. All that is needed is two AC bearings properly loaded.

A simple way is to use two washers. One for the inside race and one for the outside. Make one washer .015" thinner than the other and that will give you ABOUT the correct amount of preload.

If you look at the bearings inside race one side is thin and one side is thick. The standard way is to have the thin side face in and for the thinner shim to be on the inside race.

If you do it this way you bolt the bearings to the screw and then the OD of the bearing is clamped to the housing.

I do it different for my 704 kits but you need to cut a double sided pocket in each bearing mount. Both ways work.

[dazp1976]

Dredging up an old post because I can’t find a good definitive solution to the ball screw bearing mounts that seem to be coming back for me.

I like Arizona Video’s kits, if I didn’t already have screws I would go with their kit for the hardware.

I am curious though, are angular contact bearings needed / are they any better than a thrust washer for this application.

It would seem that there is little to no radial load. The stock thrust bearing in a decent mount may be enough?

Of course if I were needing bearings I would purchase AC bearing for this but I already have thrust bearings and can’t see it being worth it to get an AC bearing for radial load that isn’t really there….. any small amount would be dealt with the stock thrust bearing and or a good fit on a proper mount…..

I 3d printed a Y axis mount and it seems to be working well, if anyone is interested I can email the .stl files. In the long run I’ll machine the mount out of 6061, but for now it looks okay printed.

I found that if using those Chinese C7 type screws, thrust bearings don't sit on them correctly. This caused me a lot of screw wobble if remotely tightly fitted upto the screw end. Caused excessive wear on my table (X2) through this wobble.
AC bearings give a nice solid fit to the ends no matter how tight they are making the screw run true.

[toyoman]

Dredging up an old post because I can’t find a good definitive solution to the ball screw bearing mounts that seem to be coming back for me.

I like Arizona Video’s kits, if I didn’t already have screws I would go with their kit for the hardware.

I am curious though, are angular contact bearings needed / are they any better than a thrust washer for this application.

It would seem that there is little to no radial load. The stock thrust bearing in a decent mount may be enough?

Of course if I were needing bearings I would purchase AC bearing for this but I already have thrust bearings and can’t see it being worth it to get an AC bearing for radial load that isn’t really there….. any small amount would be dealt with the stock thrust bearing and or a good fit on a proper mount…..

I 3d printed a Y axis mount and it seems to be working well, if anyone is interested I can email the .stl files. In the long run I’ll machine the mount out of 6061, but for now it looks okay printed.

Angular contact bearings provide radial locationing of the screw in regards to the mount. Your standard thrust bearing does not have that ability. I have seen some weird INA-brand ball thrust bearings that are actually a combined needle bearing and thrust bearing but they are quite rare. You are correct about the low radial load on the mount bearings. AC bearings in a back-to-back or face-to-face configuration, preloaded of course, will provide plenty of resistance to axial forces and also provide alignment of the screw to the mounts. When using AC bearings in this way, your ballscrew/nut and the mounts must be properly aligned to the axis of the machine or else the components will be severely stressed.

[mactec54]

Nope, I say buggered up as akin to screwed up. Actually, I was investigating my .0035" of backlash and when I slid the table off I lifted it a bit and bent the ballscrew. My fault entirely, but now that I dialed it I know its not the ballnut, but the bearings that were the problem.

I ran into this interesting graphic. Attachment 313216

Interesting but not a good design, an open thrust bearing like this will fill with junk quite quickly, The Ac Bearing by itself is all that is needed for an average machine at the most it would have in Backlash is .0005" if the bearing is correctly installed. if you need better than the Double Ac Bearing, then you would setup (2) matched pair Ac Bearings.

[CL_MotoTech]

This is the bearing the Arizona Video kit uses, at least on X. I haven't had to change one on Y yet. Though I assume they are the same bearings in both locations. - https://www.ebay.com/itm/152624002566

There are some issues with these screws. Not so much on Y as it is short. But the X axis can be a bit tricky to get working right.

Whatever you do, make sure your bearings are well sealed. Otherwise they will just be a constant nightmare.

[dazp1976]

This is the bearing the Arizona Video kit uses, at least on X. I haven't had to change one on Y yet. Though I assume they are the same bearings in both locations. - https://www.ebay.com/itm/152624002566

There are some issues with these screws. Not so much on Y as it is short. But the X axis can be a bit tricky to get working right.

Whatever you do, make sure your bearings are well sealed. Otherwise they will just be a constant nightmare.

I found it better to make my own blocks and use 2 of 7201-2RS back to back.
Slightly thicker and larger, stronger.
Definately the way to go.
23mm thick block, 10.5mm recess either side for the bearing to slide into, 2mm central ridge for the bearings to seat.

G0704 clone with chinese C7

[CL_MotoTech]

It probably doesn't make much of a difference. The bearings aren't anywhere near their load capacity. I've had one bearing go bad in 6 years and it was clearly coolant related.

[mactec54]

AC bearings have a direction.

In the drawing I see no reason for the thrust bearings. All that is needed is two AC bearings properly loaded.

A simple way is to use two washers. One for the inside race and one for the outside. Make one washer .015" thinner than the other and that will give you ABOUT the correct amount of preload.

If you look at the bearings inside race one side is thin and one side is thick. The standard way is to have the thin side face in and for the thinner shim to be on the inside race.

If you do it this way you bolt the bearings to the screw and then the OD of the bearing is clamped to the housing.

I do it different for my 704 kits but you need to cut a double sided pocket in each bearing mount. Both ways work.

That is a terrible way to do a preload for (2) Ac Bearings, you have to measure the amount required and then, grind the right size spacer, or disassemble the Bearing and Grind the Bearings and then no spacers are needed, which is the best way to do this

The Ac Bearings the Op is using are one piece and have no why to preload the Bearings Double Row Ac Bearings are easy to use but do have a little Backlash, .0002" to .0005 which is good enough for most Hobby builds

[CL_MotoTech]

Sure, I have a home converted G0704 and when I need to replace an $8 bearing I will toss it on the surface grinder and set the preload.

Or, I could just turn a wrench and set the preload. It takes 30 seconds and the bearings last years.

[mactec54]

Sure, I have a home converted G0704 and when I need to replace an $8 bearing I will toss it on the surface grinder and set the preload.

Or, I could just turn a wrench and set the preload. It takes 30 seconds and the bearings last years.

You are obvious not serious about how your machine works, just preloading bearings like you say is a fail every time, because you are not serious about your machining you would not understand how to do it correctly, or even have a need, it does not mean that others have to follow your hack methods

[CL_MotoTech]

Please do go on. You going to call me a poo poo head next?

[dazp1976]

Please do go on. You going to call me a poo poo head next?

Reminds me of my daughter when she was 3. Called her mum that all the time.

[mactec54]

Please do go on. You going to call me a poo poo head next?

I would never have thought of something like that, but maybe it fits if you say so.

A Hack in engineering terms means, someone that is doing something that is not the norm, telling someone to do what you did is not how it should be done, no matter how it works for you, so, squeal some more