[rcazwillis]

Is is suitable to use a lovejoy connector to clamp the center piece of two bearings together to get my preload? I don't know if the single set screw in the connector will hold the pressure. My delima is the fact that I don't have any threads on the end of the ball screw. If I thread it, then I need to turn down the end for a connector to fit.

How should I proceed?

[NC Cams]

Depends on how much preload force you're trying to maintain and if you can maintain tightness of the setscrew once you do obtain the necessary preload.

If you can't assure that the tighness of the setscrew can be maintained, you already have answered your own question.

You don't have to turn the end of the shaft down for a nut. You can drill a hole in the center of the shaft and then tap it. You can then apply claming force to the end of the coupling via careful placement of a bolt into the end of the shaft.

Oh, and proceed carefully if you preloading a deep groove, non angular contact built in preloaded ball bearing....

[rcazwillis]

Good idea on the drill & tap. That will work.

The bearings I plan to use I got off Ebay. Numbers on them are 6202-2RS. I have no idea of the brand. It was listed as an angular contact bearing.

[jeffs555]

I think, a 6202 would be a deep groove bearing. The number for a metric angular contact bearing should start with a 7. This page has a good explanation of the numbering system for bearings. http://www.gizmology.net/bearings.htm

[NC Cams]

Jeffs55 if correct, your part number indicates that you have a 6202 deep groove ball bearing with 2 seals. It, per se, is NOT an angular contact bearing.

It would have to be a have a 7202 prefex to be a true, angular contact ball bearing.

Perhaps, and I do mean PERHAPS, if the bearing were custom reworked into a preloaded configuration by special grinding the face of an inner or outer ring, you would have a preloaded deep groove bearing.

This type of modification tended to be rare as they are/were only custom made for special situations when I was in the industry.

If this were the case, a reputable rework facility would etch a chevron ( >) on the bearing OD to show mounting orientation. I'd be leary of the part being properly reworked/preloaded if does not have the chevron.

EDIT: The LAST paragraph of my initial post is PARTICULARLY and signifigantly important in light of the part number and bearing design that you are trying to preload.

[RICHARD ZASTROW]

rcazwillis just a thought. 5302 or 5202 double row angular contact bearings are self preloaded and can be ordered with seals and or shields. the combined width of a pair of single groove angular contact or deep groove bearings is also wider than the width of one double groove bearing of the same id, od and ball compliment. save a bit of length.
RZ

[rcazwillis]

I seriously doubt that anything special has been done to these bearings. So, should I preload a bit or just run them as is.

[NC Cams]

To correct a popular misconception: Double row angular contact bearings (IE 5302 or 5202's) ARE NOT, I repeat and emphasize NOT internally preloaded. Manufacuring tolerances preclude the possibility of making them that way.

They do have axial and radial clearance potential (Reference NSK Technical ReportNo. E728, Topin 4.9, Page 102 dated 1992) and one affects the other.

By doing/applying an axial load and measuring the inner ring deflection, it is possible to confirm this. However, be prepared to measure in microns (0.001mm) as that is the units of measure used to confirm/quantify load versus deflection curves.

A double row bearing is capable of absorbinhg axial thrust MUCH higher than a 6202 or 6302 du to the angular contact potential of the raceways but they are TECHNICALLY NOT preloaded - not like the true preload, negative clearance as would be generated in a preloaded 7202CTYDUL, DUM or DUH bearing.

Re: use of the 6202's as is. What are you trying to do????

You can preload them with a wave washer to reduce axial slop. You can get these from McMaster Carr or perhaps Grainger. This will reduce axial slop and enhance running accuracy. This is a common practice in electric motors which is what the wave washer thingie is used for. However, as soon as you overcome the preload, you regain any axial slop you removed and instability will recur.

If you are looking for TRUE axial stiffness enhancement (like that needed for a milling spindle), you're pretty much going to have to use something like a 7202 with an appropriate built it preload or specially preload the 6202's you have..

Before the "but this is for a whatsit and I don't need an abec 7 bearing" replies start, check out the preload tables of a 7202DU( ) versus the preload potential for a wave washer copressed to the listed height.

Pick and choose as you require/truly need and make your choices based upon the numbers, not rumor, inuendo and/or sage wisdom..... If the wave washer preload suffices, fine, use it. If it won't, then you may need to spend the money for the 7200's.

The information presented in your original post is insufficient to make a 100% accurate recomendation one way or the other.....

[RICHARD ZASTROW]

NC CAMS I stand corrected. I thought I read somewhere there was a light preload that could be increased by interference fit. After checking my SKF specs. a C2 spec would be somewhere whithin 1-12 microns clearance for 5202A2RS ABEC-1
RZ

[rcazwillis]

I plan to use the bearing(s) on the Y axis of my HF 12 speed mill. I have a ball screw for the axis and the 450 in-oz motors. How do I maximize the accuracy of these bearings in this application.

Thanks for all the data so far. It is greatly appreciated.
Rick

[NC Cams]

Replace them with 7202A5TYDUHP4's

Next option is to buy some relatively inexpensive 7202BYG"s and have them preloaded with about 390N by KAF in DF config by KAF Manufacturing is in Stamford, CT

Cheapest and last option is to ask KAF if they can preload the 6202's with 50-70 lbs preload in DF configuration. You WILL have to clamp the inner and outer rings TIGHT when this is done - forget the lovejoy idea - make the space and do it right.

NOTE: the preloaded 6202's won't/can't run as smoothly or as freely as the 7202's mentioned above - long story why.

These are THE options to "maximize" your accuracy. You can kluge together any number of other fixes but these are your best alternatives.

[RICHARD ZASTROW]

NC CAMS You can get double row 60 degree angular contact with split inner races preloaded. They have thick (.984") large od (2.362") with mounting holes thru the outer race. These are designed for ball screw support. Unfortunately, they are quite expensive. INA gave me a price of $848 ea.! This is a 15mm (.591") bore p5.

[NC Cams]

As a former bearing applications engineer, I'm well aware of the co$tly bearings that can be used - HOWEVER, the DIY crowd is usually looking for inexpen$ive bearings and in many cases, downright CHEAP ones.

You can cobble up goodness knows whatever else or pay HUGE $$$'s but the ones I posted previously are the least expensive options for obtaining a TRUE precision preloaded bearing for the particular application.

[ahren]

While we're on this topic, I'd like to take this opportunity to ask NC Cams some more questions. First of all, what methodology should a DIY use to preload a bearing with a wave washer? Does this just mean selecting a washer of thickness t that will exert x N of force in y mm of deflection, and putting a spacer of thickness t-y between a nut and the bearing? Any tricks we should know about to avoid ruining a bearing?

Also, where is a good place to go to see maximum loads and maximum preloads for various different bearings? Obviously these vary by manufacturer, but are there some good rules of thumb for those of us buying things off of ebay? For example, I'd like to buy some of these, but have no idea how much I can or should preload them (or how much axial thrust they can take):

http://cgi.ebay.com/ws/eBayISAPI.dll...tem=7535858285

On a simlilar note, what's up with these?

http://cgi.ebay.com/ws/eBayISAPI.dll...tem=7594875312

They are labelled as 5203's, so they aren't really angular contact bearings, right? I would love to find some resources to help me do some of the calculations required.

Lastly NC Cams (and others), thanks for your valuable input on this topic!

Ahren

[ahren]

Umm, replying to my own post. I just re-read the part about double row angular contact bearings up above, so I guess that's what those 5203's are. However, I'm confused on these -- if they aren't internally preloaded, can they be preloaded, or do you just live with the slop when using these? And how much slop is that, typically? If it presents less than 0.001" of backlash in my system, it seems like a lot of people could live with that...

[NC Cams]

Although it is theoretically possible to "preload" a 5203 (or any ohther double row bearing) by pressing the bearing into a housing/onto a shapft, I categorically refuse to explain how to do it because I no longer have access to the critical inside geometry info and/or the clearance info needed to figure out the correct way to do it. Do it wrong and you have junk.

In some cases, 001" axial slop don't mean squat. In other cases, it is as slopy as a saddle on a sow. I don't know what you're trying to do so I can't say it will be fine or it will suck. Contact the bearing maker as axial clearance is related in a complex non-linear fashion to axial clearnace. The relation is NOT readily published and is typically another "contact the factory" value...

IN my case, 0.001 axial slop/backlsh is grossly intolearable which is why my ball screw bearings have 420 lbs preload (less than 0.0001 backlash), are ABEC 7 and cost nearly $850 each to duplicate and it would be a cold day you know where when I find the bearings on Ebay auction.

If that amount is too high of a gulp factor, you can probably cobble something together using abec 1's and wave washers or a lovejoy or even a self locking nut and some reasonable care/effort....

The L17 bearings listed as "angular contacts" on E-bay appear to be "magneto bearings". You have to look up the info/useage factors for such bearings according to the manufacturer who makes/made them. The bearing company I worked for didn't make them so I can't explain PROPERLY how much axial preload load they can take - but depending on the conact angle, it could be a lot....

As far as axial preloading of a deep groove 6203 bearing or one of similar design, 1% to 2% of the radial capacity is an acceptable "no brainer" axial preload. Raceway curvature and side finish of the groove which are NOT industry standard features govern the maximum allowable factor that can be tolerated. for any specific bearing size. As most bearing makers say in their caveat text, "conact the factory for specific recommendations when high axial loads are involved....".

As far as wave washers go, the wave washer maker provides either a load/deflection curve or a load versus deflection value. See what this is, see if it fits the 1%-2% factor, do the math to see if your numbers stack up and assemble accordingly so that you don't exceed the recommended deflection amount and you should do fine. Even if it is 3%, it should be OK. Again, doing the math is critical here - don't guess, MEASURE.

You don't want to get carried away with axial preload. Reason: the raceway gets "flatter" as you shove the ball up the side of a deep groove raceway. Thus, it is easier for the ball to brinnell the raceway the more axial thrust you apply. Or, you can crawl up over the edge of the groove with the ball thus potentially cutting the ball. As is the case in many things in life, MODERATION is the key....

Smalley is a proven source for inner and/or outer ring wave washers to preload bearings. McMaster Carr lists washers as well - I suspect hey come from Smalleyl. Check out their catalogs for loads and deflection factors and don't over compess them and you should do ok.....

[ahren]

Great! Thanks for your help and advice. In my case, I'm working on a CNC wood router, and I am looking mostly for reliability and ease of assembly. In that sense, 0.001" (heck maybe even 0.005") is perfectly fine for what I plan on doing with it. That being said, I do anticipate axial loads of 100-200 lbs on these bearings, so I'm trying to get something that will hold up under that kind of load for not a lot of $$$. However, I do enjoy learning about how to do things "right" for higher-end applications.

Given this sort of application, which understandably differs from what you are doing with your uber-bearings, can you comment on the relative robustness of a 5203 double row bearing vs a couple of cheap 7203's strapped together with a wave washer? I can certainly figure out the wave washer forces, but there is very little available literature on these shady ebay bearings. Alternatively, if you're tired of cheapskates like me asking for advice, you can also tell me to go jump in a lake . Thanks again for the info.

Ahren

[NC Cams]

L17 magneto bearings have a variable dynamic equivalent load factor (P) that changes depending on the ratio of axial force Fa versus radial force Fr applied.

First, you find Fa/Fr. You then compare it to "e" which is 0.2.

Now if Fa/Fr is < or = to e, then use the following formula:

P=1*Fr + 0*Fa

If Fa/Fr is > e, then use this formula to calc the dynamic load:

P=0.5*Fr + 2.5*Fa

Now run P thru the bearing load life calculator equations to find a rated life factor. Keep in mind that oscillatory loads don't work in most life calcs due to metal to metal contact that occurs at load reversal while oscillating.

The rated load factors for an L17 bearing are 7400N for Cr and 1500n for Cor. Grease lubed limit speed is 17000rpm and oil lubed is 20000.

I don't have a clue as to the contact angle of mag's BUT mag bearings are NOT true angular contact bearings ala a 7203 or a 5203. They are more like a deep groove ball that has one side of the outer raceway missing so you can take the thing apart axially and assemble it in special circumstance use.

If you want to eliminate chatter potential, you really want-need a preloaded bearing - period.

If you're looking for inexpensive a/c bearings, that is simply an oxymoronic wish. However, having said that, a 7203BYG is probably the cheapest one you can hope to find with an A/C bearing.

There are any number of ways to preload it (washers, shims, self locking nuts, ground preload from KAF Mfg in Stamford Ct, whatever).

For 100 to 200 lbs of axial force, you might get by with 5203's BUT with a higher chatter potential. For smooth, chatter free cuts while moving at low speed, use 7203BYG's with a medium to heavy preload....

You won't regret use of A/C's as these will spin smoothly like no 6203 could ever do because of their ability to roll along a proper raceway curvature even with ungodly amounts of preload.

5203's will accept axial loads BUT they are NOT really suited to act as true ball screw thrust bearings no matter what you may think. You can and people do misuse them but I wouldn't recommend it for reasons already cited....

If after all this, you STILL want to screw with lovejoy's, 5203's, kluged up 6203's, waves or whatever other cheap a$$ alternative, I"m sure you can find a nearby lake.... Luckily your's don't freeze over like ours do up north of M/D line....

[RICHARD ZASTROW]

NC CAMS You would know better than I, but wouldn't one of those "new fangled" split inner ring 4-point contact bearings be a reasonably priced alternative for a compromise on the low force ball screw? This assumes the correct sizees are available.

[NC Cams]

Call around for pricing from bearing distributors, compare and then let us know what you find.

I'd be surprised if you find cheaper alternatives for name brand bearings from legitimate distributors (NOT clearing house auction prices from Ebay) than what I've proposed....

Nobody ever made a 4 point ANYTHING that was inexpensively priced and/or readilly available....