Angular Contact Bearing Spacing?

[ngr1]

Hello Everyone, I am about to convert a round column mill to CNC and I was hoping to gather some advice regarding the assembly of my bearing blocks.

I have reviewed Bob Warfields site (Thanks Bob!) regarding the appropriate bearing configuration, but I am still unclear as to whether or not I need to space my angular contact bearings.

Below is my current drawing of the ballscrew ends and a parts list.

Bearing # 7203 – 40 deg. Contact angle, dual-Load Angular-Contact Steel Ball Bearings—ABEC-3, from http://www.mcmaster.com/


Nook Rolled Ballscrew Stock- .631 x .2 Pitch, Single Nut-Preloaded (apparently backlash is typically .0005" or better) http://www.homeshopcnc.com/page3.html

Any help would be appreciated.

Thanks… Norman

[ViperTX]

Well you can't go wrong with a spacer....I always use spacers....well Misumi does....and I use their bearing blocks.....take a look at their catalog, it's online at misumiusa.com....I went with them because they are cost effective.

The spacers only support the outer periphery of the bearing and they should be precision spacers.

[ngr1]

Thanks for the tip ViperTX.

But my problem still remains about which spacer to select or whether or not a spacer is even required considering the inexpensive bearings I am using.

Norman

[philbur]

I think in this application the only purpose of the outer spacer is to ensure that the inner races do not contact each other when you apply preload. So as long as thespacer is truely square then a millimeter or two would be more than enough for thickness.

Regards
Phil

Originally Posted by ngr1 Thanks for the tip ViperTX. But my problem still remains about which spacer to select or whether or not a spacer is even required considering the inexpensive bearings I am using. Norman

[ViperTX]

philbur....has it correct.

The problem when you preload the bearings....is that the inner bearing center section is pushed towards the outer bearings center by the the ballscrew or leadscrew bearing stop surface.....the outer bearings center is being pushed towards the inner bearings center by the preload nut.....so, it is possible.....check the dimensions of your bearings....that they may not preload at all.....

[NC Cams]

Philbur has it partially correct and Vipertx brings it home.

Ball screw bearings are usually mounted DF so as to allow for more self alignment potential. Has to do with the theoretical intersection point of the contact angles. Do NOT expect off the shelf bearings to have preload when you get them. Thus, buying what you'll buy and bolting them in could result in slop beyond belief. Dah WHAT????

If you mount the bearings right next to each other with preload, you have to do some work. You first have to axially load one of the bearings and then measure the inner raceway "axial offset" ( both bearings).

THen you either grind off the ofsetting amount EVENLY to maintain raceway to side squareness or find a way to space out the outer race ring to generate the same preload induced raceway axial offset.

If you space the bearings appart with inner and outer ring spacers, you only have to measure the bearing preload offset. Then you simply make the spacers with different inner an outer ring axial thicknesses INTENTIONALLY so as to affect the same preload when you clamp them together.

See how simple it is????

[philbur]

Accurately measuring the bearing offset and making inner and outer spacers to the necessary matching tolerances is not everybody’s cup of tea. So preload can and is very often achieve with only an outer spacer, with preload adjustment then achieved by torqueing the retaining nut, which is then backed up by a locknut. I believe this method is much easier to achieve and is perfectly acceptable for this particular application.

Regards
Phil



spacers

Originally Posted by NC Cams Philbur has it partially correct and Vipertx brings it home. Ball screw bearings are usually mounted DF so as to allow for more self alignment potential. Has to do with the theoretical intersection point of the contact angles. Do NOT expect off the shelf bearings to have preload when you get them. Thus, buying what you'll buy and bolting them in could result in slop beyond belief. Dah WHAT???? If you mount the bearings right next to each other with preload, you have to do some work. You first have to axially load one of the bearings and then measure the inner raceway "axial offset" ( both bearings). THen you either grind off the ofsetting amount EVENLY to maintain raceway to side squareness or find a way to space out the outer race ring to generate the same preload induced raceway axial offset. If you space the bearings appart with inner and outer ring spacers, you only have to measure the bearing preload offset. Then you simply make the spacers with different inner an outer ring axial thicknesses INTENTIONALLY so as to affect the same preload when you clamp them together. See how simple it is????

[NC Cams]

Unless someone KNOWS or figures out how many turns of the lock nut yields so as to generate however much preload, you're in the same boat.

Namely, guessing what the preload is by variable position preload"feel" of a bearing as you tighten it. This is NOT necessarily in the better interest of the bearings' static load capability if you OOPS and go too far. Nor will it necessarily result in proper preload or optimum functionality - for a lite/moderate preload, it probably won't matter.

BUT, if you're looking for a STIFF, deflection free ball screw with a true A/C bearing fitted thereto, you're talking preloads (depending on size and contact angle) that can range from 60 to over 260lbs preload PER BEARING. My 60 deg true ball screw bearings are fitted at over 500lb TOTAL preload that turn glass smooth and with LESS rotating torque than kluged up 6204's preloaded wtih only 75lbs preload TOTAl.

Having done preloading work and calculations professionally, I've seen where many "guesstimates" at preload offset (as in "that feels about right" ), even in a clinical setting, resulted in poor bearing stiffness (what's being sought) and/or less than optimum service life (what guesstimates often yield).

Even though I knew better, I tried it the spacer/nut turn method and the results were far less than even moderately stellar. This is why I spent the money for professionally preloaded bearings.

Yes, DIY'ing of bearing raceway offset is NOT for the ill equipped or the faint of heart and even less appropriate for anything but patient individuals. However, there is a "right" way to preload A/C bearings and a "that should work way" that the user can opt for - the raceway offset measurement procedure under a known preload value is simply how the pro's do it.

The DIY"er can cut corners as they see fit (torque turn or other methods) if and do it any way they choose. However, they must be willing to live with whatever results they generate for themselves - good, bad or indifferent.

BTDT (torque turn and even shimming t'ween the bearings) and simply wasn't satisfied with the compromise (poor stiffness and high turning force). Went the pro route and was NOT disappointed.

Simply put, caveat emptor....

[philbur]

Hi NC CamS,

I would not dare to argue with your technical knowledge on the subject. However the problem is that measuring and machining to a precision of <0.0002" or purchasing 3 match pairs of angular contact bearings costing more that the milling machine they will be installed on is beyond 95% of the members. I think they and I would benefit much more from more doable solutions based on your professional experience.

Regards
Phil

Originally Posted by NC Cams Unless someone KNOWS or figures out how many turns of the lock nut yields so as to generate however much preload, you're in the same boat. Namely, guessing what the preload is by variable position preload"feel" of a bearing as you tighten it. This is NOT necessarily in the better interest of the bearings' static load capability if you OOPS and go too far. Nor will it necessarily result in proper preload or optimum functionality - for a lite/moderate preload, it probably won't matter. BUT, if you're looking for a STIFF, deflection free ball screw with a true A/C bearing fitted thereto, you're talking preloads (depending on size and contact angle) that can range from 60 to over 260lbs preload PER BEARING. My 60 deg true ball screw bearings are fitted at over 500lb TOTAL preload that turn glass smooth and with LESS rotating torque than kluged up 6204's preloaded wtih only 75lbs preload TOTAl. Having done preloading work and calculations professionally, I've seen where many "guesstimates" at preload offset (as in "that feels about right" ), even in a clinical setting, resulted in poor bearing stiffness (what's being sought) and/or less than optimum service life (what guesstimates often yield). Even though I knew better, I tried it the spacer/nut turn method and the results were far less than even moderately stellar. This is why I spent the money for professionally preloaded bearings. Yes, DIY'ing of bearing raceway offset is NOT for the ill equipped or the faint of heart and even less appropriate for anything but patient individuals. However, there is a "right" way to preload A/C bearings and a "that should work way" that the user can opt for - the raceway offset measurement procedure under a known preload value is simply how the pro's do it. The DIY"er can cut corners as they see fit (torque turn or other methods) if and do it any way they choose. However, they must be willing to live with whatever results they generate for themselves - good, bad or indifferent. BTDT (torque turn and even shimming t'ween the bearings) and simply wasn't satisfied with the compromise (poor stiffness and high turning force). Went the pro route and was NOT disappointed. Simply put, caveat emptor....

[NC Cams]

In an economy of words, even though I know how to professionally preload bearings, I made an effort to use DIY methods to change the bearing preload for the "ball screw bearings" (actually modified 6204's) that were factory installed by the OEM in my milling machine. What a nightmare.

For those who have not read my Eztrak ball screw "escapade", the Cliff notes version was that we used thin shims to shim the outer raceways appart to generate more preload.

Reason: to take out direction change backlash/bearing deflection. As any DIY'er might do, we looked long and hard for consistant albeit thin shims - we settled on and ultimately used aluminum foil (Reynolds brand). Besides, it was a Sunday afternoon and that's all we had.

You have no idea how hard it is/was to perfectly cut out MULTIPLE thin washers to shim only the other ring so as to add preload to the factory ground preloade 6204 bearings.

We then when thru an iterative loop to add shims, measure backlash & turn around flex and then redo it with another shim variation.

Anyone who's changed ball screw bearings on a mill table with a CNC drive attached to the same end KNOWS how much of a god awful PITA this effort was.

The results: the more shims you added, the less deflection you got BUT with more and more turning resistance.

Even when we shimmed to the point where it took a strong arm on a breaker bar to turn the handle, we still had nearly 0.0005" direction change flats.

Moreover, keeping the foil from tearing and keeping it smooth/flat was critical because if it got "lumpy", it was like turning a ratcheting ball screw with fits and starts.

We spent nearly a week screwing with that masturbation of time - not to mention thousands of dollars of lost productivity from my shop business - but I learned none the less.

Thus, if you want an inexplensive solution, cut some shims from aluminum foil and shim away - don't be surprised with regard a learning experience that you should expect.

Once we learned WHAT NOT TO DO (We thought we knew how to do it) we then tried to duplicate it in a less costly way. You see, our major customer went another direction and we lost maybe 65% of our business at the time so COST was a real concern - we had time but no $$$. Anyway, weeks later, after MULTIPLE low dollar attempts, we gave up on the kluged up 6204's and looked into the use of true A/C's. Friends helped us score bearings that cost what a small mill costs - but comparable performing ones can be bought on Ebay or via wise shopping on a budget but you have to look.

The thing about A/C's is that they are ALL more expensive than deep groove ball bearings (DGBB) - hence, you're gonna pay more no matter what. The choice becomes that of:


HOW MUCH $ DO YOU SPEND?

or

How much TIME do you want to spend experimenting (and perhaps ruining bearings if you go a bit too far)?

In light of the thousands' of dollars of man hours in time that we spent figuring out that kluging up 6204's only to find out that shimming does NOT work effectively with DGBB's, we had some surplus true ball screw bearings properly preloaded by a pro and quit screwing around.

Consequently, when this thread started, I decided to offer the professional "how to" and let it be after that. We'll simply will let others reinvent the wheel via shimming and kluging - BTDT and it quite simply sucked.

Now, if you simply bolt together some 7204 @40 deg contact angle, DO NOT EXPECT them to be preloaded. There will probably be axial end play - and lots of it.

THis is why they make wave washer shims for motors - to affect a constant pressure preload via a "spring". Trouble is, this isn't enough to adequately bi-directionally preload a bearing for a ball screw.

The reason why I'm against "torque turn" preloading is simple: unless you REALLY know HOW to do it, there can be a real high possibility of easily brinnelling the bearing which ruins them - period & instantly.

A teeny tiny amount of over-reefing and POOF, the bearings has divots permanently pressed into the raceways. Sorry, but telling someone on a M/B "how much to reef" is simply not doable. Professional advice, take it or leave it.

The "unequal shim/spacer" is not THAT difficult to do. If you have or can get some tubing, you can cut spacers and either mill them the same height or surface grind them to the same height - sanding on glass plate to make them perfectly flat/smooth/parallel - easy DIY stuff.

Once they are equal thickness, you can use the foil trick to find out what sort of thickness differential best suits your needs. Simple, effective and also both suitable for DIY or pro use.

Don't think so?

IT is a little known fact that the machine tool makers do NOT always use preloaded duplex bearings. Some of them have use NON-preloaded bearings and then install unequal spacers between the inner and outer rings to affect and/or adjust the preload for the individual machine. HINT: be careful if you replace bearings with a factory spacer inbetween, they may have unequal thicknesses and "hidden preload factor" built in.

How do they know how much washer thickness differential?

They dead load the bearings with weights and measure the bearing raceway offset under load to get what they want.

How do they find out what you want? Install a bearing and keep adding preload until they get the stiffness they want (then duplicate it) or do an OOPS and toast a bearing during the process.

That's when they do something different (a tad less preload) or live with as good as they can get.

BTW, you simply can NOT calculate the preload offset because NOBODY gives out the true raceway dimensions and tolerances needed for anyone to calculate the load/offset relationship. Factory preloaded bearings are ALL offset loaded and measured and ground to suit.

YES, anyone can do torque turn preloading - even I tried to do it. But I learned after lots of trying that it isn't viable for some bearing variations or it can suck if you try using it on the wrong bearing design.

My professional experiences and my personal kluging are reflected in my original post and this treatise. These experiences led me to point out the prescribed way to affect preload establishment on a ball screw bearing pair.

The reader can now choose between shimming, torque turning or buying preloaded bearings or any permutation of methods that suits their fancy or pocket book.

After what I went thru, however, I'd still recommend buying something good - when you buy the best you can't afford, you generally are NEVER disappointed.

BTW: my machine tool tech thought my "high dollar" fix was way too unaffordable for even his commercial customers. Even after going thru the time versus bearing cost analysis, he still couldn't justify the cost.

But, hot damn, he loved how the ball screw on my machine (which he'd tuned hundreds of for the OEM over the years) was so deadly accurate.

Guess who called last week to sheepishly ask for the bearing P/N's so he could bolt in bearings instead of spending hours trying to "tune" backlash out of a ball screw fitted mills anymore???

You now have viable options based upon "expertise and experience". Pick the one that best suits your budget and/or performance needs or expectations.

[ngr1]

Thanks for the info gentlemen.

I cannot pretend to understand everything that has been suggested but I will try a precision spacer to support the outer periphery of the bearings.

Norman

[NC Cams]

NGR1 = try to understand the prior post - if you're going to use spacers, you MAY or MAY NOT want equal spacer widths between inner and outer rings. This depend on whether or not your core bearings are or are NOT preloaded already.

By making the spacers unequal widths, you can affect a preload. It is important, however, to note that the unequality is expressed in MICRONS (0.001mm).

If you're going to do DIY preloading, you NEED to understand HOW TO DO IT and HOW MUCH is appropriate.

Yes, it is not rocket science but it is easier to damage bearings via improperly preloading them than some people think...