angular contact bearings, or not

[klick0]

Ok, i'm trying to determine how to eliminate backlash from the standpoint of the servo or stepper motor. After reading the ballscrew basics a few times, and finding the section on thrust loads, it mentions angular contact bearings as a way to eliminate this thrust backlash that could occur in your driving motor. Which I fully understand and want to eliminate, but after searching for suppliers of angular contact bearings, i'm not finding much that makes sense to me.

mcmaster has angular contact bearings with just look like normal old bearings to me, I understand the whole offset part that if you get 2 then make your own block, but i'd rather buy an assembly already pre-made. So if going through mcmaster, i'm assuming I would buy 2 of these per ball screw and then have to make custom bearing assemblies to hold these properly.

I understand the need to eliminate the thrust load being distributed to my servo motor, and first off, is my comprehension that angular contact bearings are the key? If they are, where is a good supplier of pre-made blocks that would support this. I'm assuming the block would need to attach itself to the ballscrew through some kind of set screw (a series of screws i would assume), but have seen no such setup.

The ballscrew basics post shows one supplier I believe but everything is in millimeters, and www.roton.com (where I was planning on getting ballscrews from) is all in inches of course. Now I know I could roughly translate these, but I attempted that and found nothing that lined up properly. I'm planning on using 3/4" ballscrews for all axises, and need thrust/angular contact bearings inbetween the ballscrew and my servo motors, well, inline with them, as to eliminate any backlash my servo's may cause thrust wise.

Also, on a side note, i've read many good posts about roton's balllscrews, but do people think they are good screws? I'm going with HiWin 15MM linear slides (heavy duty line) for all axis's. I'm planning using roton's ballscrews, but they don't post much for the accuracies of their screws and ballnuts. Their prices please me, but if the quality isn't there i'll have to look elsewhere.

Thanks,
Ross

[HuFlungDung]

I snooped around the Roton site, and see a conspicuous lack of information about the accuracies of their ballscrews, but the price tells me that they are rolled thread, and probably come "backlash free", which means there is lots of backlash which they do not charge extra for

I suspect you'll need to have mounts for dual angular contact bearings custom made. The shaft mount requires more than setscrews, rather, it will require a turned shoulder on the screw and a threaded nut to capture the pair of bearings on the screw. The housing needs to be of such construction that the bearings are also firmly captured, ie, no press fits or snap retaining rings because these are unreliable end location methods.

[NC Cams]

WE chased the backlash deal all over hell and back (look for the Extrak thread that 'explains it) and it can be done.

A/C's take radial and axial thrus and vector sum it in two directions. Axis and radial thrust. The contact angle you choose depends on what you want/need in the way of axial versus radial thrust capability.

15 deg bearings are used more for radial loading as the majority of the thrust vecotr is assubmed radially. As your contatct angle gets higher, your axial thrust AND axial stiffness capability get HIGHER as well. This is why ball screw supports have high contact angles of almost 60 deg - high azial load absorbtion and high axial stiffness.

Then you get into the issue of preload. THe more preload you have, the STIFFEr the bearing. WE chased preload versus contact angle all over hell as well. If you want zero compliance, buy true BALL SCREW SUPPORT bearings and prelod the hell ouf them. These are special made, high contact angle and HEAVILY preloaded. Ours spin like glass smooth, They have drag but not impossible to turn and the accuracy is DEAD ON with nearly 500lbs of preload.

The Extrak thread has the bearing P/N's for the "good" part numbers. You can come up with lower cost alteranteives using 40 deg contact angle A/C's and heavy position preload. True TAC style ball screw bearings are 60 deg and PRICEY but deadly stiff and accruate. Loose/sloppy supports are a waste of time so spend some money and mount the stuff well in solid billet.

Custom bearing preloading is not easy to do DIY. We tried and it is a PITA. Get help from a pro. KAF MFg in Stamford CT does it. I cunno about cost amymore but it is worth the cost. He did mine and I paid like anybody else.

[Oldmanandhistoy]

Hi,
While we have NC Cams attention; could I jump in with a quick question?

Is there anything wrong with using a couple of thrust bearings and one or two radial bearings in place of A/C bearings?

Thanks,
John

[NC Cams]

It is usually esaier to package an integrated A/C than to stack up both a radial and an axial thrust combo.

I"ve also found that a pure thrust bearing (as in a radial needle) is a PITA to prelload. I never learned why but suffice it to say that rolling ball A/C's have lower rotating friction than needle thrusts. I think it has to do with the friction factor that a few large balls with point contatct have as compared to a lot of small rollers have with line contact.

In a word, it is mostly packaging - simply pakaging.

[Oldmanandhistoy]

It is usually esaier to package an integrated A/C than to stack up both a radial and an axial thrust combo.

I"ve also found that a pure thrust bearing (as in a radial needle) is a PITA to prelload. I never learned why but suffice it to say that rolling ball A/C's have lower rotating friction than needle thrusts. I think it has to do with the friction factor that a few large balls with point contatct have as compared to a lot of small rollers have with line contact.

In a word, it is mostly packaging - simply pakaging.

I was thinking using ball thrust bearing.

The reason I ask is that finding 6mm bore A/C’s is not easy if not impossible but 6mm id thrust and radial bearings is not a problem.

Thank you,
John

[lgalla]

The McMaster ballscrews are C7 or .004/ft
Roton's are Tor C10 .009/ft.Apparently you can request .003.
Check http://www.misumiusa.com/ They will machine the ends to your specs.
Larry

[Oldmanandhistoy]

I
The reason I ask is that finding 6mm bore A/C’s is not easy if not impossible but 6mm id thrust and radial bearings is not a problem.

I take it back; I have just found some 4mm,5mm and 6mm bore

[ger21]

Do you have a link for those?

[Oldmanandhistoy]

Do you have a link for those?

http://www.arceurotrade.co.uk/en-gb/dept_21.html

[HuFlungDung]

A flat needle roller thrust bearing is destined for failure under constant load situations. The angular velocity of the races differs across the width of the race, but the needle rollers have a constant circumferential velocity across their length. Something is slipping somewhere and its got to cause friction if the length of the contact area is significantly different (like on small diameter bearings using long rollers).

I've never seen a thrust bearing with tapered rollers, but theoretically, there is likely a set of parameters that would allow the tapered roller to 'naturally roll' around the circle with minimal slippage. But then the damn things would need to be contained, so you'd have end thrust dragging you down

Ball bearings also have the same problem with slippage of the balls in a full contact raceway, and this is why they have particular contact angles designed in. Either define the contact points in the design, or the bearing wears its own clearance. This is why stuffing oversize balls into a worn ballscrew does not produce optimum preload and/or running life.

[NC Cams]

For small diameters, you can often times and literally axially load deep groove ball bearngs and let/shove the balls up against the sides of the raceway - in effect creating an A/C - You won''t necessarily be in a honed portion of the raceway nor have the smoothest operation but it will be a small diameter A/C bearing in effect.

WE did this triick on some low cost overhead fans and the client was satisfied even th0ogh the net result was clearly and effectively a kluged up bearing - it worked and it lived. game over.

Can you axially load ball bearings? yes. do you want to? NO because you can't accurately predict the life due to the level of unknowns (IE lack of full raceway support, lack of full ball contct, etc). Even so, kluged up stuff can and does work, sometimes with astounding results.